E.B.R.O. 497-01 THE ONTARIO ENERGY BOARD IN THE MATTER OF the Ontario Energy Board Act, R.S.O. 1990, c. O.13; AND IN THE MATTER OF an Application by The Consumers' Gas Company Ltd. for approval of an incentive mechanism in relation to the Operation and Maintenance Expense component of its cost of service, effective during the 2000 through 2002 fiscal years, and an incentive mechanism in relation to Demand Side Management. Hearing held at 2300 Yonge Street, 25th Floor, Hearing Room No. 1, Toronto, Ontario on Thursday, February 4, 1999 commencing at 9:07 a.m. ---------- VOLUME 7 ---------- B E F O R E : H. GAIL MORRISON PRESIDING MEMBER PAUL VLAHOS Member ROGER M. HIGGIN Member 971 A P P E A R A N C E S JENNIFER LEA ) Board Technical Staff EDWARD SWEET ) KATHI LITT ) FRED CASS Enbridge Consumers Gas, Applicant MURRAY KLIPPENSTEIN Pollution Probe ROBERT WARREN Consumers Association of Canada (CAC) TOM BRETT Metropolitan Separate School Board, Ontario Association of School Board Officials and CAESCO BETH SYMES ) Alliance of Manufacturers CAROL STREET ) and Exporters, Canada GUY PRATTE ) Industrial Gas Users PETER C.P. THOMPSON, Q.C.) Association (IGUA) DAVID POCH Green Energy Coalition IAN MONDROW HVAC Coalition ELIZABETH DeMARCO Coalition for Efficient Energy Distribution (CEED) THOMAS ADAMS Energy Probe PHILIPPA LAWSON Ontario Coalition Against MICHAEL JANIGAN Poverty (OCAP) MICHAEL MORRISON Ontario Association of Physical Plant Administrators LINDA ANDERSON Union Gas Preliminary Matters 972 ---Upon commencing at 9:07 a.m. THE PRESIDING MEMBER: Please be seated. Good morning. MR. CASS: Good morning. THE PRESIDING MEMBER: Any preliminary matters? Mr. Cass? MR. CASS: Just in a very minor way, Madam Chair, I wanted to report that undertaking response H4.4 I believe was passed around late yesterday, actually, and undertaking response H4.3 has been passed around this morning. I'm told the others are all in progress and we're trying very hard to get them completed today, but the one that may present some difficulty is 4.2, but we're going to try to get them all completed and filed today. THE PRESIDING MEMBER: Thank you, Mr. Cass. Remind me. Is 4.2 the one on the affiliate transactions? MR. CASS: "In year" effects. THE PRESIDING MEMBER: Oh, okay. Thank you. Now, Mr. Janigan, are you prepared to proceed here? MR. JANIGAN: Yes, I am, Madam Chair. THE PRESIDING MEMBER: Go ahead. MR. JANIGAN: I would like to call John Randolph Norsworthy as a witness for CAC, IGUA and OCAP. You have to step forward to be sworn. JOHN RANDOLPH NORSWORTHY, Ph.D.; Sworn. Norsworthy 973 dr ex (Janigan) DIRECT EXAMINATION BY MR. JANIGAN: Q. Dr. Norsworthy, first with respect to your qualifications, you received your BA and Ph.D. in Economics from the University of Virginia? A. Yes. Q. I believe you did postdoctoral studies in Law and Economics at University of Chicago? A. Mm-hmm. Q. You have appended to your Curriculum Vitae which is attached to your evidence a list of books and publications primarily, but not exclusively on the subject of productivity? A. Yes. Q. And previous to your position of Professor at Rensselaer Polytechnic Institute, you were employed as a Senior Research Fellow and Chief Consulting Economist with the U.S. Bureau of the Census? A. Yes. Q. And you founded their Center for Economic Studies? A. Yes. Q. And I believe for nine years, you were Chief of the Productivity Research Division at the U.S. Bureau of Labor Statistics? A. Correct. Q. And as I have indicated, you are currently Professor of Economics and Finance at Rensselaer Polytechnic Institute? Norsworthy 974 dr ex (Janigan) A. Mm-hmm. Q. And Co-Director of the Center for Finance and Technology? A. Yes. Q. And your CV also indicates that you have testified in a number of regulatory proceedings upon the subject of productivity. Is that correct? A. Yes, it is. Q. I understand, from the list that is in your CV, that you have testified in regulatory commissions on behalf, first of all, of industry, on behalf of AT&T in the price caps proceeding in 1994 in front of FCC. Is that correct? A. Yes. Q. And you have testified for various regulatory commissions on the subject of productivity? A. Yes, mm-hmm. Q. And you have also brought testimony for consumer groups; in particular, productivity analysis involved in the price caps proceeding in the CRTC in 1996? Is that correct? A. Yes, mm-hmm. MR. JANIGAN: I would like to ask that Dr. Norsworthy be admitted as an expert for productivity analysis. THE PRESIDING MEMBER: The Board is prepared to accept him, Mr. Janigan. MR. JANIGAN: Q. Now, Dr. Norsworthy, you are Norsworthy 975 dr ex (Janigan) responsible for the evidence that was filed in your name and for the interrogatory responses that you have filed on January 26th? A. Yes. Q. And these were prepared by you and under your supervision? A. Yes, mm-hmm. Q. Now, are there any corrections to your evidence? A. Yes, there are. At page 4, to make certain that the correction is in what you have, the fifth line from the end of the first paragraph showing that the regulator really did underestimate the company's potential productivity gains. Q. My copy shows "underestimated" currently-- A. Yes. Q. --and I believe that correction was made at our office, but we just wanted to make sure that it had been. A. Yes. That's an important distinction, for sure. At page 10, following equation No. 2, the second point there, the change in TFP sub TE, it should read; not NE. Q. I wonder if you could give the full sentence, so that we make sure that -- A. Yes. Delta TFP sub TE is the rate of TFP growth in the inflation benchmark entity. And then adding Norsworthy 976 dr ex (Janigan) 'for example, the total economy'. The following line is deleted. IPRE is the input price, inflation, the regulated company. Then page 22, the fifth line from the bottom of the paragraph on page 22, the word "services" is repeated. The sentence should read: 'It is also reasonable to expect that these additional services could have been delivered at less than average cost.' The first occurrence of "services" should be deleted. At page 23, the next-to-last line: 'The company was doing the opposite.' Insert "was" between "company" and "doing". Its capital efficiency was declining. At page 29, point D, the sentence: 'This incentive will provide additional challenges to the Board to protect the company's customers from being overcharged.' Insert "being". Change "overcharging" to "overcharged" for the services purchased from the unregulated affiliates. That's all, I believe. Q. And with these corrections, is your evidence accurate, to the best of your knowledge? A. Yes, it is. Q. Now, I want to deal in a preliminary way with your evidence and I don't want to do any review of the principal points that you have covered, but I just want to touch upon several areas, particularly some areas that Norsworthy 977 dr ex (Janigan) have been in dispute thus far in this proceeding. A. Mm-hmm. Q. First of all, I want to deal with your observations concerning productivity in the company and the industry. And how do you measure productivity? How did you measure it in this case? A. Yes. We want to measure productivity in order to appraise the company's performance, its overall performance and the accepted vehicle for doing that is what's called total factor productivity. In order to measure total factor productivity, we need to have measures of the expenses that are associated with each major category of inputs purchased by the company; in this instance, capital and labour and materials and purchase services. I will say something about gas in just a minute. Now, for the output, we need to have the gas volumes and revenues by rate category. In addition to the historical expenditures, we also have to have either the prices or the quantities of resources for each of the categories. This information was requested from the company and the labour information was provided. Briefly concerning capital, we requested investment and depreciation and 1987 or some other year benchmark by category; however, the benchmark measures and the depreciation measures were not provided. We were told that they were unavailable. Norsworthy 978 dr ex (Janigan) Now, concerning gas, because gas costs are mainly pass-through in Canada and because the structure is set up that brokers may provide -- may purchase and sell gas to individual companies, it is reasonable to exclude gas from the productivity measure. In principle, however, to the extent that the company buys gas to serve its customers, there is less incentive to bargain sharply and, in principle, it might be better to have the gas that is purchased under the productivity umbrella. However, most of the analysis that I have done is based on the exclusion of gas, so that we are talking essentially about a total factor or multi-factor productivity measure that excludes gas. Now, in order to appraise the company's performance in a TFP measure, it was necessary to use available information on the capital input because we could not get what was required to construct a more conventional measure based on the perpetual inventory method; however, I will describe why this is not by any means a terrible thing as long as it's used properly. Materials is a very important component of overall non-gas costs - about 14 per cent in 1997 - and the costs of materials has been rising fairly substantially approximately since 1987. Based on the data provided by Statistics Canada on purchases of materials and services inputs by the Canadian gas industry, I constructed a price index for materials. Norsworthy 979 dr ex (Janigan) Its growth differs rather substantially from that of the Ontario CPI. It grows more slowly by about eight-tenths of a per cent a year. Compared to using the Ontario CPI instead of the materials price index, the Ontario CPI overstates inflation in materials and this, between 1987 and 1997, results in an overstatement of about $6.1-million. Since materials' input has grown this effect is greater in '97 than it was in '87. Now, the capital input is very important. It's about 55 per cent of non-gas costs in 1997. It's easy to calculate the capital charge to customers. It's what's left over from revenues after labour and O&M expenses are removed. The calculation of the quantity of capital input is really the problem. The usual method, as I mentioned, is the perpetual inventory method which accounts for depreciation and changes in the prices of investment goods and which depends upon a benchmark for each component of the capital input in some common year. The benchmark is typically taken as the net book value of capital in a particular year. Now, the alternative is to use the rate base which is, in fact, the net book value of capital. It is net of depreciation as determined by the regulator. The allowed return on that rate base is determined by the regulator in such a way as to allow for capital recovery and the reasonable return on equity after Norsworthy 980 dr ex (Janigan) taxes and the cost of debt. The fact that the rate base increased substantially at ECG is not really at issue. That is, indeed, a fact. That rate base affects -- reflects additions to capital, new investment and deductions from capital, depreciation and retirements at values judged by the Board to be appropriate for the company to recover its capital investment. Q. Dr. Norsworthy, I want to stop you right there. A. Please. Q. You covered two main points that were raised in the evidence-in-chief and I believe under cross-examination by Dr. Fuss. I just want, because these points were contained in a lot of other material, I just want to make sure that we have those points. The first point that you made was with respect to the use of your materials price index and, in particular, why it was to be used and what the difference between using that is and the Ontario CPI. A. Mm-hmm. Q. Could you, please, briefly review that point? A. Yes. A consumer price index generally has -- generally grows faster than prices for inputs made by industry. This is largely because the intervening factors include the retail margin and productivity growth, and retailing is not typically sufficiently fast to -- not as fast as it is in a manufacturing industry. Norsworthy 981 dr ex (Janigan) And so it turns out that if we accurately measure the price index for the materials input based on information from Statistics Canada, we come up with a rate of growth which is about eight-tenths of a per cent per year slower than the rate of growth of the Ontario CPI. Q. Now, Dr. Norsworthy, just to interrupt you here. A. Yes. Q. Dr. Fuss indicates that this difference doesn't matter in terms of the construction of your performance-based measure. A. Yes. I believe what Dr. Fuss is referring to is the notion that in an "X" factor regime, if you reduce the materials price index, then that has the effect of raising input and lowering total factor productivity on the one hand, and it has an offsetting effect on the input price differential on the other hand. This is entirely correct. However, from the perspective of appraising the company's performance, you need to look at the best estimate of total factor productivity that you can get; namely, one in which the inputs are adequately deflated and the output is properly measured. Q. Now, I want to move you ahead to your use of rate base numbers for the purpose of calculation of capital. A. Yes. Q. And I believe you've reviewed the transcript Norsworthy 982 dr ex (Janigan) of February the 2nd on the testimony of Dr. Fuss. Dr. Fuss has indicated that had this is, perhaps to paraphrase, an invalid calculation and presents a flaw in your analysis. How do you respond to that? A. Well, first of all, if we had the information to calculate a perpetual inventory measure of capital input we would use that as well as the rate-based calculation that I have provided. However, we didn't have that information. And Fuss asks, I guess rhetorically, whether the Board would permit Consumers Gas to operate with capital intensity 50 per cent higher than the industry. This question really confuses the growth of capital input with the level of capital input; that is, capital intensity refers to how much capital there is per unit of output or capital per customer or whatever. What we find is that the capital cost at Consumers Gas has risen faster than that in the industry, has risen faster and not necessarily at a higher level. I do not have the information to look at that, as I indicated. So I make no assertions that the levels of capital input in ECG and in the industry are related in any particular way. Essentially, Dr. Fuss is dealing with the problem of his own construction that grows out of his misinterpretation of levels and rates of growth. Q. Okay. I just want to briefly review that. Norsworthy 983 dr ex (Janigan) You indicated you did not have the data to do the perpetual inventory method of capital calculations? A. Right. We did request it. Q. Now, is this lack of data fatal to your analysis? A. Not at all. Q. What is the reason for that? A. There are two reasons. First of all, as I indicated, the rate base is, in fact, adjusted by depreciation and rewarded by capital charges which are sufficient to guarantee capital recovery. The rate of depreciation that's determined by the regulatory authority essentially is based on the notion that capital should be taken out of the rate base when it's no longer in service and depreciated over its useful life. Of course, retirements from the capital stock should be taken out directly. So we have in the rate base a measure of the net capital input which is directly associated with the allowed depreciation and which correspondingly is the basis for the allowed return to capital. One of the reasons why economists doing productivity studies typically use an alternative construct is, in fact, that they do not have the financial information that would be required to construct an alternative. The perpetual inventory method does not take into account, for example, differences between debt and equity. Norsworthy 984 dr ex (Janigan) It does not -- in other words, it does not take account of the capital structure of the enterprise and it does not take account of other factors that are commonly considered both in regulatory hearings and in industry financial analysis generally. So I think the best approach is to use both measures where possible, but certainly it can't be argued, given the regulatory environment, that the rate base is an invalid measure of capital input. Q. Doctor, I want to deal briefly with the observations concerning the company's proposed MFP regime and would like you to comment briefly on your observations concerning the design. A. Yes. The O&M/MFP that is offered by the company, first of all, uses customers as a measure of output. This is really a very unusual practice. Normally when less than the total inputs are used to measure productivity, the conventional measure is still the best available measure of output; output per labour hour, output per employee, et cetera. And that is not the case in the proposed O&M multi-factor productivity measure. Moreover, the input is -- as I indicated, the materials input component of O&M is inadequately deflated, specifically, it is over deflated, which results in an upward bias of the O&M productivity measure. The proper design would be to use a total factor productivity measure that includes capital and labour and Norsworthy 985 dr ex (Janigan) other purchase inputs as well. If one limits one's vision - and there are reasons why this works to the detriment of Consumers - but, if one limits one's vision to an O&M measure of productivity, then the appropriate measure of output is still the volumes of gas weighted by their unit revenues for the different rate classes; and the appropriate measure of input is, of course, the best achievable based on proper deflation of labour and materials. I certainly have no quarrel with the deflation of labour as it's carried out in the company's measure. Q. Now, very briefly, Dr. Norsworthy, I want to deal with some specific points that came up in the testimony of Dr. Fuss. Dr. Fuss explained on transcript reference 713 that he used a level productivity comparison to determine the cause of ECG's relatively poor productivity performance over the 1987 to '92 period. Do you have any comment on that approach? A. Yes. He used a level productivity comparison, which presumably is based on customers as a measure of output and materials, deflated by the Ontario CPI. That's the first point. And concerning customers as a measure of output, the charges to customers are based on the volumes of gas consumed by the customers. The access charges which may be regarded -- or access costs which may be regarded as fixed by the rate-making process also has to be, of Norsworthy 986 dr ex (Janigan) course, supplemented by the other costs of providing gas services. And that's why the rates for gas are measured according to the volume of gas consumed and the revenues are collected on the basis of the volumes of gas consumed. So the design of the productivity, as I indicated, the O&M productivity is flawed. Now, that applies whether a productivity level measure is used or whether a productivity growth rate is calculated. There will be, in the level measure, the same kind of error. In order to illustrate this very briefly. Suppose that we have the gas network in place, then when more gas is purchased by a customer, the flow of services to that customer is greater and the revenues collected from that customer is greater. This is the general rationale for rate making in network-type industries. Under the company's proposed plan, however, the economies of scale that come about when more gas is used per customer will accrue -- the benefits of those economies of scale will accrue directly to the company and will not be reflected in the O&M-type productivity measure proposed by Dr. Fuss. Finally, it seems to me it would be difficult to do a level comparison based on data from Statistics Canada because Statistics Canada told me that they did not, in fact, have data on gas customers, they had data only on gas volumes and associated rates. Norsworthy 987 dr ex (Janigan) So this level comparison of total factor productivity is -- first of all, cannot be constructed exclusively on the basis of data for the industry from Statistics Canada and is flawed for the reasons that I indicated, because both the output and the inputs are mismeasured. Q. And finally, one last question, Dr. Norsworthy. The Averch Johnson effect has been discussed in this proceeding, and Dr. Fuss indicated in his evidence-in-chief that, in fact, the Averch Johnson effect is a dead issue. Could you comment upon that, please? A. Yes. Dr. Fuss is directing his comments to the initial version of the Averch Johnson paper and, to that extent, he is correct; the original version of the paper was a very limited theoretical model, it looked only at capital and labour and looked at it in a static, single period analysis. Dr. Joskow's assertion, cited by Dr. Fuss, that uncertainty about validation of capital expenditure could conceivably lead the company to under invest in capital if it doubted the validation of that capital expenditure by the regulatory authority. This is the narrow framework of the original AJ paper and there's no real quarrel with that. But, more broadly, the Averch Johnson effect is also a term that's used - and, admittedly it's used loosely - in regulatory analysis to describe the real and Norsworthy 988 dr ex (Janigan) current concern that a regulated company, with a guaranteed rate of return to its capital, has the incentive to substitute capital where cost recovery is guaranteed for other inputs; not only labour, but materials, purchase services, et cetera, where future prices are uncertain. However we may name this phenomenon, the real fact of over investment in capital has to be a real concern for the performance of Consumers Gas because, we observe, that the capital charge per customer has increased about 40 per cent in the 1987-1994 period at the company, as compared to virtually no increase in the industry at large. Q. Thank you, Dr. Norsworthy. MR. JANIGAN: Thank you, Panel, for your indulgence. Those are my questions in-chief. THE PRESIDING MEMBER: Thank you, Mr. Janigan. Which counsel are interested in questioning Dr. Norsworthy? Mr. Cass and Board Staff? Would you like to go ahead then, Ms. Lea? MS. LEA: Thank you. CROSS-EXAMINATION BY MS. LEA: Q. Dr. Norsworthy, in your evidence-in-chief you answered quite a few of my questions with respect to productivity, but I do have other questions as well and a few follow-ups on what you said. Norsworthy 989 cr-ex (Lea) A. Mm-hmm. Q. I wonder if I could begin with a question out of the introductory part of your testimony. It's at page 5 in the first partial paragraph there. You indicate that: 'Under appropriate incentives, regulated companies regularly find ways to improve efficiency without beggering the quality of service.' Is it your view that such appropriate incentives are embodied in this plan that we have before us? A. No. Q. What incentives could be included to make this happen? A. First of all, I think incentives based on a broader measure of performance are required. Under present circumstances, or under the proposed PBR there is an even stronger incentive to substitute capital for O&M inputs than is present under conventional cost of service regulation. So the first step would be to broaden the performance measure to include capital as well as O&M inputs. Q. If broadening the proposed performance-based plan was not an option, if that -- or if that option were rejected, is there anything else that a regulator can do to, first, provide appropriate incentives; and, secondly, guard against this substitution of capital inputs for Norsworthy 990 cr-ex (Lea) labour or other O&M inputs here? A. Well, I think the first thing that one would want to do is to adjust the O&M measure so that it, in fact, measures output per unit of O&M input with the proper measure of output and the proper measure of real materials inputs; that is, deflated materials inputs. I think, also - and I guess there have been some recent changes in exactly how affiliate relations are to be treated. In the presence of affiliate... I'm sorry. In the presence of unregulated affiliates, if the company is permitted to buy inputs from the unregulated affiliate, then the charges that are made to the company may be inflated. And even though this results in a small loss based on the O&M performance measure, there can be a more than compensating increase in the profitability of the overall company; that is, the holding company that embraces both the regulated company and the affiliate. This is not a pipe dream. This has turned out to be a very substantial problem in the local telephone industry in the U.S. where several companies have established unregulated affiliates and then wound up cream-skimming the labour force in the regulated company; that is, hiring the best workers, and charging more for services than the costs of those services when they were provided internally. There is additional flexibility when one considers that the relationship with the affiliates goes on through time, so that even if initially the prices Norsworthy 991 cr-ex (Lea) charged for the outsourced services are not substantially out of line from what they were when they were provided within the company, there is the danger and the incentive to raise the charge for those services and thus provide profits at the expense of the customers. Q. Do you see this, the need to guard against these dangers, the ones that you have outlined, as increasing the regulatory burden that this Board will have to use in regulating Consumers Gas because of the nature of this proposed plan? A. Yes. The regulatory burden would be reduced if, for example, all inputs were embraced within the performance plan, and the regulatory burden would also be reduced if there were not the alternative of purchasing inputs from unregulated affiliates. Q. In the absence of those things, do you think the regulatory burden may increase in the circumstance that we face if we accept the company's proposal as filed? A. Yes. Q. You have indicated in your evidence that labour inputs, wages and prices and materials should be -- that input should be based on market-based data. As you're aware, in this hearing the company is putting forward company-specific historic data for those purposes. Is it ever appropriate to use company-specific data for these inputs? A. When you're appraising the performance of the Norsworthy 992 cr-ex (Lea) company, of course, you have to use company-specific information in order to measure the costs and the productivity performance of the company. But in judging the performance of the company, then you need the basis of comparison from outside the company because essentially the company is a monopoly of the services within its service area. Q. There was one follow-up question to your discussion of affiliate transactions. Can we just look for a moment at page 16 of your evidence, sir? There was one thing that we didn't understand in Table 2 in the fifth column. That's entitled 'Loss to...' -- page 16. A. Mm-hmm. Q. Fifth column. 'Loss to regulated company on incentive plan equals 0.76 per cent times O&M increase.' What is the source of the 0.76 per cent? We didn't understand where that came from. A. Yes. This derives from the 0.6 or 0.63 proposed by the company. I raised that somewhat to be within what might have been the range of contemplated stretch factors, and so this somewhat overstates the loss that the company would undergo if the 0.63 or 0.6 factor were applied. Q. Thank you. One moment. ---Off the record discussion. MS. LEA: Q. I wonder if we could look at page 19. I don't know if you need to turn it up or not. There Norsworthy 993 cr-ex (Lea) was just another quote I wanted to investigate of your evidence. On page 19 in the first full sentence there, you talk about the company having under-performed the industry by about 2 per cent per year. A. Mm-hmm. Q. And this, I understand it, was based on total factor productivity? A. With gas included, yes. Q. With gas, okay. A. Yes. Q. From this statement, can we draw any conclusions about multi-factor productivity from this observation about total factor productivity? A. Well, since the multi-factor productivity is measured directly as part of the preparation of Table 3, I think it can be said that the gap associated with TFP including gas, when we adjust that by taking gas out, when we adjust it by taking capital out, and when we properly measure output based on how customers pay for it, and properly deflate inputs based on the prices that prevail in the industry, the prices and patterns of expenditures that prevail in the industry, then we come up with a multi-factor productivity measure for O&M that is consistent with this 2 per cent. And one could, in fact, parse it out piece by piece, starting with the 2 per cent and moving to the other measures of productivity, terminating with the Norsworthy 994 cr-ex (Lea) multi-factor productivity measure of negative fourth-tenths of a per cent per year. And then one can make the transition by backing out the output and input measurement corrections and arrive at the measure suggested by the company. Q. Okay. Thank you. That's helpful. There was also a statement you made on the next page of your evidence, page 20, in the first paragraph, the last sentence. You indicated there that on your evaluation, the performance gap was 5.7 per cent for the time period that you have considered. And you indicated also that that overstates the problem. I wonder if you could expand a little bit on that statement, please? A. Yes. '95 was the last year for which data was available for the industry on a revised basis. It was not a particularly good year for the company and so it may be that that down-tick in company productivity that year makes it a little bit less appropriate. I think in some of the other discussions; for example, in Table 6, for precisely that reason, I terminate the measurement comparing the company and the industry in 1994 instead of 1995. So that's what I refer to when I say that the problem may be somewhat overstated. Q. Thank you. All right. I think we understand that now. Norsworthy 995 cr-ex (Lea) In explaining why it selected number of customers as its output factor, one of the things that the company said in answer to an interrogatory was that when it used volumetric-based productivity estimates, it ended up with negative productivity. And in its judgment, this demonstrated the inappropriateness of using a volumetric factor because the company believed its productivity to have been positive. I wonder if you could comment on the exercise of judgment by the company as I have paraphrased it? A. Yes. The measure for the industry is based on volumetric measures of output and, in fact, includes what is going on at Consumers Gas, as well. The industry shows positive productivity growth. I can understand that the company may feel that there are elements in the services it provides that may be not be reflected in the volumetric measure, but that nonetheless is the standard way to measure output for productivity analysis. One of the reasons, as I indicated, is that customers are charged not on a flat rate basis, but on the basis of the volumes of gas that they consume. And the prices or the rates that are set per unit for those volumes of gas do, in fact, take into account the various costs, including the connection charge. The company has had a poor productivity record in recent years. I think that is unequivocal. If you look at Table 3, you will, of course, see that by whatever Norsworthy 996 cr-ex (Lea) measure of total factor productivity or multi-factor productivity, the company's productivity has declined. Q. Dr. Norsworthy, one thing that we're grappling with is if the productivity factor included in the formula is lowered from what the company proposes, that, in the way that the proposed formula works, as I understand it, increases in fact the O&M budget on the left side of the equation. Do you believe that would be an appropriate result; that the Board should consider lowering the productivity factor and thereby increasing the O&M budget? A. No. I think the appropriate correction or fix is to include capital, as well, so that these artificial incentives and, indeed, misleading measures are mitigated. Q. If, because of the nature of the application before it, the Board found that it could not import capital into this equation at this time, what in your view would be an appropriate and fair productivity factor to use for the company if the formula has to remain as proposed? A. If the formula must remain as proposed, then the output measure should be corrected, as I indicated, to reflect the volumetric measure and the input measure should be corrected to reflect the prices paid for O&M inputs by the company. The best estimate of those prices is based on the industry materials price index. Norsworthy 997 cr-ex (Lea) Q. Thank you. Turning to another section of your evidence, I think at page 33 of the evidence, you talk about using sharing mechanisms and you indicate that a sharing approach gives teeth to the performance targets. Are there other "teeth" which we can apply to performance targets? The company opposes the imposition of a sharing mechanism. Can you comment on whether there are any other mechanisms that might achieve the same result? A. There are other mechanisms that might achieve the same result, for example, more substantially rewarding higher performance. If you sort of look at the sharing, one of the approaches that has been used in some other jurisdictions is to set a single high target for the company's performance and if the company meets that relatively high target, then it is permitted to keep all of the additional profits. This, of course, is predicated on the idea that the company -- I'm sorry, that the price at the end of a particular performance period will be ratcheted down to reflect the productivity gains that the company has accrued. Q. Thank you. If this Board was concerned about making sure that ratepayers benefit as much as possible from productivity gains, would a sharing mechanism be the most effective way to achieve that, or is the company's Norsworthy 998 cr-ex (Lea) proposal sufficient here? A. No, I think the company's proposal is insufficient. Sharing is one way to put -- to extend the dental analogy, "teeth" in the proposal, and in particular you'll notice that in table 11. Q. At page, sir? A. Page 33. Q. Thank you. A. And 34. Q. Thank you. A. You'll notice that this is set up so that there are three steps in the sharing mechanism. Now, the more steps, the more regulatory burden on the one hand; but the more steps, the more the greater the incremental incentive to the company to improve its performance and to pass along benefits to the customers. It may be worth while to mention that in the interstate access regulation, the Federal Communications Commission initially had a fairly moderate target for the price caps and a three- or four-step sharing arrangement. Five out of the seven companies elected the highest arrangement despite their statements that they in other -- their other statements that they will lose money by doing so. And, in fact, one of the companies that elected the lowest rate wanted to change to the highest rate after the first year. In the second round the same thing was observed even though the sharing requirements has been ratcheted Norsworthy 999 cr-ex (Lea) up. Again, the company elected the highest -- five of the seven companies elected the highest productivity target and met that target. So this is the basis for my enthusiasm for teeth in the performance -- productivity performance plan. Q. Thank you, sir. My last question relates to monitoring of a PRB plan. At page 44 of your evidence, you discuss the dangers and difficulties of using data from the company. You indicate in the last paragraph there, the second paragraph: 'The companies will have the incentive to misreport the data, to overstate the cost of providing regulated services and to invest in capital computer software and other inputs that support production of both regulated and unregulated services.' Sir, what advice would you have for a regulator who is about to embark on a PRB measure such as the one before us in terms of making sure it gets sufficient data to guard against these difficulties? A. Unfortunately, the obvious way to deal with it is somewhat more energetic monitoring, more regulatory burden both on the company and on the Board. Q. Thank you. One moment. ---Off the record discussion MS. LEA: Q. Sir, can you give us some examples, practical examples, of how this has been achieved by regulators in the past? Norsworthy 1000 cr-ex (Lea) We're interested in understanding the best way to monitor a report on this -- monitor this type of scheme. A. The Federal Communications Commission set forth a productivity measurement scheme which was, in fact, based predominantly on this so-called performance-based productivity measurement suggestion that I developed with AT&T. That specified in terms of quantities already reported exactly how productivity must be calculated for the companies and reported on an annual basis. That was within the framework of reporting that had already been established by the FCC based on the fairly highly evolved regulatory system -- regulatory accounting system that came about through the National Association of Regulated Utility Commissioners which, of course, also consulted with the Financial Standards Accounting Board and so forth -- Financial Accounting Standards Board, I'm sorry. The difficulty is that in addition to requiring this information to be reported, it is probably more important after the imposition of such a plan to audit what is reported. When I mention misreporting here and in other places, the notion is not that the company is malevolent. Only that it will tell its story in such a way that reflects the interests of the company. So-called incentive regulation is established to address exactly that, to recognize that the company will Norsworthy 1001 cr-ex (Lea) act in its own interest and to try to harness those actions, the action the company takes into its own interest, at least partially to the benefits of the customers. Q. Thank you. In considering then what will be necessary in terms of receiving data and auditing it, this Board may have several choices with respect to this application. One of the choices it probably does not have at this time or at least it has been suggested this choice is not available is to order the company to bring forward a comprehensive PRB plan immediately. Given what you said about regulatory burden, how should the Board evaluate whether it should go ahead with a targeted PRB man or reject the plan and wait for something more comprehensive? Before you answer that question, I wonder if you could comment on whether you think the regulatory burden from a comprehensive plan would be significantly greater than the regulatory burden this Board will assume if it accepts the targeted plan put forward. A. The regulatory burden of a comprehensive plan I would expect would be less than the regulatory burden of this particular plan because in order to enforce this particular plan to make sure that it's operating in the interest of customers you are going to have to more closely monitor the expenditures on capital and out-sourcing to unregulated affiliates. Norsworthy 1002 cr-ex (Lea) With the more comprehensive plan, it's easier to look at the total amount of revenue that's collected and the total expenses required to generate those revenues than to break it down into pieces, some of which must be done anyway, but you don't have to worry about the details of substituting capital for O&M inputs in the same way that you do when, in fact, you create an incentive to substitute capital, a further incentive to substitute capital for O&M inputs. Q. Thank you. So then if this Board is to evaluate this application, how should it go about assessing the benefits of having some incentive mechanism applied to this company against the disbenefit of increased regulatory burden? A. Well, the regulatory burden occurs in one quarter and the benefits to customers in another. And the theoretical answer to that is sort of embraced in an analysis that includes the customers and the company's interest more or less on a dollar-for-dollar basis. That is the theoretical approach. The more realistic approach is that, in fact, under the proposed plan I think, with all respect to the originators thereof and so forth, the customers are going to be worse off irrespective of regulatory burden. The regulatory burden is going to be greater to reduce the damage to customers that might otherwise occur. Q. How will the customers be worse off, sir? A. The customers would be worse off because the Norsworthy 1003 cr-ex (Lea) uneconomic substitution of capital for other inputs would take place and in the presence of the unregulated affiliate the door is open. Q. Okay. Thank you. MS. LEA: Thank you very much, sir, for your answers. Thank you, Madam Chair. Those are my questions. THE PRESIDING MEMBER: Thank you, Ms. Lea. As we said yesterday, we will be sitting until noon and then back again at two. So I think we will take our morning break now and we'll take 20 minutes, please. ---Recessed at 10:15 a.m. ---On resuming at 10:50 a.m. THE PRESIDING MEMBER: Please be seated. MS. LEA: Mr. Cass, the things that are being handed out now, are they something to be used in cross-examination or -- I didn't know whether you wanted exhibit numbers now or not? MR. CASS: It appears to me that what is being handed out now is a copy of the Draft Code which was intended to be an attachment to Undertaking Response H3.4. MS. LEA: Okay, thank you. THE PRESIDING MEMBER: Thank you. Mr. Cass, would you like to go ahead? MR. CASS: Thank you, Madam Chair. CROSS-EXAMINATION BY MR. CASS: Q. Dr. Norsworthy, if I may I'd like to start just by trying to have an understanding of the process Norsworthy 1004 cr-ex (Cass) that you went through in completing your work. Now, correct me if I'm oversimplifying, but would it be fair to say that the process would be one of accumulating data, analyzing the data and then reaching conclusions? A. That certainly is a procedural summary, but I sort of started off by thinking about what was going on and reading parts of the company's plan and so forth. Q. All right, thank you. And I'll probably come back to parts of that, but it's really the procedure I'm just talking about at this point in time. A. Mm-hmm. Q. Now, in terms of the accumulation of data, I think you've already talked in your evidence today about what you did to get information from the company. What did you do to get industry information? A. For the materials price index deflation, I was in touch with Statistics Canada's Input/Output Division and they provided information on expenditures by the industry by year in current and in constant dollars. For the productivity information, I was in touch with the Productivity Division at Stat Canada and from them I got, in the first round, measures of inputs in various stages of aggregation: capital, labour, energy, materials, services and an output measure, and later I got supplementary information. The first set of information covered the period from '61 to '92 and, subsequently, there was more Norsworthy 1005 cr-ex (Cass) information on cost shares for these various inputs and, as well, the information for labour and capital extended to '95. Q. All right. Now, you've referred to the Input/Output Division and the Productivity Division. Are those the only divisions of Statistics Canada that you consulted? A. I don't recall whether I thrashed around a little bit before I got there, but they certainly are the only two divisions that provided information to me. Q. I see. Well, you had mentioned in some of your testimony today that you were unable to get information about customer numbers. A. That's right. Q. And I had wondered whether you'd inquired of the Energy Statistics Group of Statistics Canada for that information. And I take it that you did not? A. That's correct. Q. Thank you. And I gather from the material that you also considered the possibility of data from the United States, but you concluded that the data was not appropriate because of the vertically integrated nature of the industry? A. Yes. Mm-hmm. Q. Okay. So in terms of productivity and industry numbers, it was the Statistics Canada data that you relied upon; correct? A. Yes. Norsworthy 1006 cr-ex (Cass) Q. Now, I also gather from the covering letter that accompanied your evidence that there were some difficulties with Statistics Canada in you getting that data. And I wonder if you could elaborate on the difficulties for me, please? A. Well, I'm not sure I would describe it as difficulties with Statistics Canada. Stat Canada was, and as far as I know is, in the process of revising and updating their productivity data and, in fact, the first time I called they didn't have the information for the gas industry that they subsequently sent to me, including capital and labour input measures and cost shares and so forth going to 1995. Partly, I think, there was a mutual exploration of: 'Gee, what exactly do you mean?' and 'What do you have?', 'What do you want?', and so forth. And so there was, if you like - in economics we use the term "tatonnement", there was a kind of a groping, but we finally got to what could be called a usable set of data. Now, in that -- well, go ahead. Q. All right. Just a couple of follow-up questions. You said Statistics Canada is in the process of revising and updating the productivity data. So that is an ongoing process now, even as we speak; is it? A. I believe so, yes. Norsworthy 1007 cr-ex (Cass) Q. And then you also said you got to a usable set of data? A. Yes. Q. How did you do that? A. What was required in order to fill out the whole program of calculations was, first of all, to get cost shares for the energy, materials, capital, labour, services through 1995. Second, to get quantities of capital, labour and the other inputs through to 1995. In order to do that, Stat Canada provided the capital and labor information which, of course, is the bulk of the costs in the Canadian gas industry, but it was necessary, and I concurred in their advice, that the best thing to do was to extend the price series from the earlier period to the years '93 to '95. And, of course, I had developed a deflator for materials inputs. They had a deflator for services, and at this point we still were talking about the industry and the costs and inputs in the industry exclusive of gas. Then, in order to be able to make a comparison, which also included the gas input, which is almost two-thirds of the total cost of the company - although it's less in recent years, I guess two-thirds is an average over the period that I looked at, what I had to do is essentially take the proportion of gas costs to total costs -- or to other costs for Enbridge and apply that in order to get a gas cost for the Canadian gas industry. And then I used a deflator to get at the gas volume Norsworthy 1008 cr-ex (Cass) deflator from Enbridge. Now, the fact is that using the Enbridge data to supplement the industry-level data sort of establishes a little bit of a bias in favour of Enbridge; that is, to the extent that the technologies of production differ and we use the Enbridge weights, that is -- that makes the result come out to be a little bit closer to the performance that would be expected if the industry had the same gas cost share as Enbridge. Q. But that's a conventional way of doing it; is it not, Dr. Norsworthy? A. Yes. Q. Thank you. Now, in terms of this industry data that you obtained from Statistics Canada, did you do anything to inform yourself as to what companies are included in that data? A. I think a fair answer is no. Q. And were you able to form any conclusions as to the comparability of the inputs in that data to the inputs in the data you obtained from Enbridge Consumers Gas? A. Yes. My understanding is that the same concepts are applied in determining the expenditures for labour and for capital and for materials and services combined and, of course, gas is the same in each case. Q. And how did you form that understanding? A. From the discussion in Enbridge's submission about, for example, the comparability of materials in the Norsworthy 1009 cr-ex (Cass) industry and in the company and so forth. Q. Okay. So we can look to the evidence in this case in that regard then? That is what you were looking at? A. I think so, yes. Q. Thank you. A. I looked at what evidence? I'm not sure I understood. Q. What you were looking to when you formed that understanding was the evidence in this case, I take it? A. Yes, uh-huh. Q. Thank you. And I just want to put a time frame around this. When were you retained in connection with this matter? A. I think the initial discussion was late summer. Q. Did you have a formal retainer at some point in time? A. Yes. I guess, if you mean formal retainer in terms of the undertaking that I would do the work, I guess that came, what, September, I'm not -- of course, most of September is summer. Q. Do you have a letter of engagement? A. Yes. I didn't bring it. Q. I see. Okay. And when were your discussions with Statistics Canada? A. I believe the initial discussions were some time in October, prior to going to Statistics Canada, Norsworthy 1010 cr-ex (Cass) because in my view Statistics Canada's methods and the quality of their data are better than those of the U.S. I attacked the problem of getting data from the Bureau of Labour Statistics and Bureau of Economic Analysis in the U.S. first, and when it proved that those data would be inadequate, I, at that point, went to Stat Canada. Q. Now, having obtained this information from Statistics Canada, what analysis did you do of the Canadian gas industry in order to form a conclusion about any adjustments that you ought to make to the data? A. Predominantly I looked at the approximate size of Consumers gas relative to the rest of the industry, and I also looked at the size of BC Gas compared to the rest of the industry. And beyond that, with respect to making adjustments such as weatherization and so forth, I did not pursue the matter because, in fact, in a performance-based measure, weatherization or adjustment for differential loads caused by weather differences, that adjustment is really not appropriate for the sort of productivity measure that we were trying to establish. In particular, I noted earlier, for example, that 1995 was not a very good year for the industry and I chose not to use that as the terminal year for some of the analysis. That sort of thing, I looked at; that is, the trends in the variables that I had data for. But I did not undertake, for example, a long-term historical Norsworthy 1011 cr-ex (Cass) analysis of the industry, although I did compute productivity for the industry based on the Stat Canada information back to '61. Q. Okay. So what you've told me about then would be the extent of your analysis of the Canadian gas industry for this purpose, I take it? A. I think that's right. I consulted... Let's see. I guess Gas Facts, which is a publication in the U.S., has some information about the Canadian gas industry and there were some little things like that. Q. Okay. A. But you have pretty well summarized it. Q. Thank you. And no criticism whatsoever intended, but is this your first specific project with exposure to the Canadian gas industry? A. For practical purposes, absolutely. Professor Fuss and I start dead even on that. Q. I just want to run something by you, Dr. Norsworthy, that I believe would be apparent, but I want to be sure you agree with it. A. Mm-hmm. Q. Would I be correct in thinking that your conclusions depend on the quality of your analysis and which, in turn, would depend on the quality of your data? A. Among other things. But the quality of the data and the trends in the data that come out of analyzing the data both in terms of productivity and trends in the individual series that go into the measurement of Norsworthy 1012 cr-ex (Cass) productivity, that go into the description of what's happening in the industry. For example, the charges to customers and how that tracks with gas volumes, with the number of customers, and so forth; all of this enters in as well as a thorough understanding of productivity measurement and of the regulatory environment, the decision environment of the company and so forth. Q. All right. A. I -- Q. Sorry. A. Yes. I will refrain from carrying on. [Laughter] Q. All right. I appreciate your elaboration on the proposition I put to you, but I take it you would agree that to the extent that there is any question about the data or any question about appropriate adjustments that have not been made to the data, then that would undermine the analysis and the conclusions which flow therefrom; correct? A. No. I wouldn't really agree with that. It would depend on what those questions were, the context in which the questions were raised, and the purposes for which the questions were raised. For example, I would not agree that simply raising a question about weatherization or simply raising a question about the capital measure is sufficient to undermine the data analysis and the conclusions based on Norsworthy 1013 cr-ex (Cass) the data. It's not enough just to ask a question. The question has to be answered in the context of the study that I undertook. Q. All right. And perhaps we can leave that part of it to argument. But to the extent that the questions which are raised are legitimate ones and can be shown to be well-founded, then the proposition I put to you is correct; right? A. I would not agree because the questions may be legitimate and well-founded and appropriately answered, and under those circumstances the data and the conclusions based on the data are not undermined. Q. All right. I think we're in agreement on that. Thank you, Dr. Norsworthy. A. All right. Sure. Q. I want to take you to a particular number, but for that purpose I think you should turn up the exhibit, just in fairness to you. All I want to do is get one number from it. And I've had someone open the binder for you further down the table. It's Exhibit D, section 4.44. A. Mm-hmm. Q. And this was an interrogatory to Enbridge Consumers Gas which asked for production of a study by authors named Lowry and Kaufmann? A. Yes. Q. The name of the study is "Forecasting the Norsworthy 1014 cr-ex (Cass) Productivity Growth of Natural Gas Distributors". Do you have that in front of you? A. Mm-hmm. Q. If I could just ask you to turn up page 8, so that we can look at the number that I want to ask you about, Dr. Norsworthy? Page 8 of Exhibit D, section 4.44. A. Yes. Q. Now, on that page you will see Table 2, which deals with the average annual growth rate in productivity, national LDCs, 1984 to '93. This, of course, is referring to LDCs in the United States. And under the productivity column, we see a reference to 0.4 as the average annual growth rate. Do you see that? A. Mm-hmm. Q. Now, I will take you to your evidence on this, but you would agree with me that that growth rate is significantly lower than that which you found for the Canadian industry? A. Yes. I would want to know, of course, whether this, how this measure of productivity was constructed, whether it includes gas and there's no doubt, for example, that output in the U.S. in the gas industry is growing somewhat more slowly than in Canada and very often by no means invariably output growth and productivity growth tend to go along together. Q. All right. Well, if we turn up Table 3 of your evidence at page 18, Dr. Norsworthy, looking to the second-last line of your table, the Canadian industry TFP, Norsworthy 1015 cr-ex (Cass) total factor productivity, no gas that you arrived at for a slightly different time period was 1.66 per cent? A. Mm-hmm. Q. Now, I take it that you haven't done any analysis, Dr. Norsworthy, to determine why you have come up with that figure for the Canadian industry that appears to be larger, significantly larger than what Messrs. Lowry and Kaufmann came up with for the American industry? A. Well, I think probably, I would have to look at the procedures that Lowry and Kaufmann undertook in detail, but -- Q. Fair enough. But my question, Dr. Norsworthy, was just to ask, to confirm that you haven't done that. You haven't done any reconciliation between your figures and figures that might be produced from data in the United States? You haven't done that analysis; is that correct? A. That's correct. Q. All right. Thank you. And, excuse me, still at the same page of your evidence, Dr. Norsworthy, if I can just get myself back there, in the fourth line up from the bottom on page 18, you indicate: 'Detailed input data for the industry were available only up to 1992.' A. Mm-hmm. Q. So that's consistent with what you have told us about the data that you were able to obtain from Norsworthy 1016 cr-ex (Cass) Statistics Canada? A. Yes, mm-hmm. Q. All right. Now can I ask you to flip over to page 25 of your report, Dr. Norsworthy? A. Mm-hmm. Q. In the second line under the heading 'The Proposed Output Measure' on page 25, there's an indication: 'Increases in sales per customer, a major source of economies of scale.' Now, I won't beat around the bush about this, Dr. Norsworthy. I'll just put it to you directly. I would suggest to you that in the context of the gas industry, particularly in Ontario, it is unrealistic to think about economies of scale being achieved from increases in sales per customer. Is that not correct? A. I think that characterizes the period for which I have data for Consumers Gas. My understanding of how the investment program of the company is designed is that there is the anticipation that at some point the ratio of benefits to costs will come into balance. I guess there must be some such anticipation in the case of Consumers Gas in order to have justified the capital expansion which is evident, whether one looks only at the growth of the rate base or whether one simply looks at the charges for capital. Now, in order for that to be expected to come Norsworthy 1017 cr-ex (Cass) into balance in the future, there must be some anticipation of increased density, increased gas deliveries per customer. If there are such increases, then clearly that would lead to economies of scale which, as I indicate, would be captured by the company. Q. Well, let's look a little further down the page, Dr. Norsworthy-- A. Mm-hmm. Q. --and just talk a little more about some of the things you have referred to. In the second sentence of the second paragraph, you refer to the fact that: 'The number of customers has grown at a rate of 3.85 per cent per year while unweighted gas volume has increased only 2.7 per cent.' A. Yes. Q. That, in fact, is revealing a diseconomy of scale insofar as volume is concerned, isn't it? A. Well, a diseconomy of density, yes. Q. All right. And then you go on to give two reasons for what you call the disparity in the growth rates, still on page 25. A. Oh, sorry. Mm-hmm. Q. Okay. And you give two reasons. The first is: 'Services (plural) per customer have declined.' That's a concept that seems a little foreign to me and I just wondered, first of all, is that a Norsworthy 1018 cr-ex (Cass) terminology that one might expect to hear in the telecommunications business, "declining services per customer" or "increasing services per customer"? A. Sure. Q. And what does it mean? A. In many service-based industries, that terminology is used. Q. And what would it mean, for example, in the context of the telecommunications industry? A. If we take the services provided, for example, by the local telephone companies, there are three broad categories of services. And the practice blessed by the Federal Communication Commission is to aggregate those categories of services using their unit revenues. And this is the consequent, or this leads to the measure of output for the industry. So that is what is meant by the catch-all term "services" in local telephone companies. Now, those are the regulated services. There are other unregulated services that are provided that would also be embraced by the term "services" in general, but would not be included in a measure of regulated company productivity. Q. Can you just quickly give me some examples of what you would be talking about here in the telecommunications context in terms of these services per customer? A. Oh, sure. Number of calls per customer, Norsworthy 1019 cr-ex (Cass) number of minutes of either local or long distance time per customer, number of special access lines which can be put on a customer basis and so forth. In general, there is a whole complex of services which includes measured services and -- go ahead. Please proceed. Q. Yes, thank you. The second reason you've given, Dr. Norsworthy, for what you've call a diseconomy of density is the mix of services in the weighted total has shifted toward lower valued services. Would I right in thinking, again, that that's a concept that has some real meaning in the tele- communications business? A. Well, it also has real meaning in the gas business. Q. Well, let's start first with the telecommunication basis. What does it mean there? A. Suppose, for example, that we have a shift from long distance services, which have a higher tariff, to local services which typically have a much lower tariff, on a minutes-of-use basis, then that shift -- now, factually the shift has been in the other direction, but that shift would in the telecom industry cause a decline in the measured service based on a change in the mix of services. Q. Now, I would like to talk for a minute about the Ontario gas industry, Dr. Norsworthy. What I'd like to suggest to you is that what Norsworthy 1020 cr-ex (Cass) you've called a disparity in growth rates is conservation of gas in Ontario, correct? A. I don't know about the specific contributions of gas conservation in Ontario. I think if conservation were applied in such a way that it affected all the of the inputs, then there would be no necessity to expand the capital inputs involved in the delivery of gas services and one would also be able to economize other inputs such as labour and materials and purchase services and so forth. Q. Well, I'd like to suggest to you, Dr. Norsworthy, that two reasons for what you've called diseconomies of density are, number one, the increased energy efficiency of the building stock; and number two, increased energy efficiency of the gas-burning appliances; isn't that correct? A. I will provisionally accept -- 'accept' your assertion. Are these improvements not also carried out elsewhere in the Canadian gas industry? Q. Well, I actually have a question or two for you on that, but, first of all, you haven't analyzed this phenomenon that I'm talking about at this point in time, have you, this conservation effect? A. No, I haven't. Q. Right. A. Nor I believe was there -- well, go ahead. Q. You haven't made any adjustments in your analysis for disproportionate conservation effects in Norsworthy 1021 cr-ex (Cass) different utilities across Canada, correct? A. I'am not sure that such an adjustment would be appropriate. Q. Well -- A. When one measures productivity, one measures output per unit of purchased input. That is the first step. There may be subsequent steps where one attempts to parse out the causes of changes in output per unit of total factor input. Q. Right. And hopefully I can avoid arguing with you here on the stand. We can leave that for later. I just wanted to confirm, you haven't made any adjustment for what I just referred to, for the possibility of disproportionate conservation effects? A. I'm not sure what disproportionate conservation effects means. Q. All right. But you haven't made any adjustment -- Let me put it to you this way: Are you aware of the Board's decision in this jurisdiction in EBO 169? I take it not? A. That's correct. Q. And you haven't made, accordingly, any adjustment for the impacts of demand side management arising out of that decision? A. That is correct, but I wouldn't want to be interpreted as acknowledging that such adjustment was appropriate. Norsworthy 1022 cr-ex (Cass) Q. That's fine. I take it from your evidence, Dr. Norsworthy, that you have experience in productivity analysis that cuts across a number of industries and you gave us a biographical stret -- sketch at page 656 of your evidence. I'm sorry, stretch... [laughter] ... that refers to -- A. I am old, but I don't know whether we should call it a "biographical stretch" or not. Q. I apologize. A biographical sketch that identifies the industries where you've had some focus in terms of your productivity analysis? A. It indicates some of the industries. Q. Yes. A. I guess it certainly doesn't indicate all of them. Q. All right. But the ones you chose to identify in that biographical sketch I assume are the ones that you determine in your own mind were worthy of being highlighted there? A. Well, since I wrote the biographical sketch the biography has stretched to include a book that will be published this year on the airline industry, the U.S. airline industry, which is another network-based industry. Q. Of course. Actually, I thought airlines were referred to there, but in any event -- A. Okay. That's fine. Q. In your curriculum vitae you give us some of your experience in the regulatory area, and I gather from Norsworthy 1023 cr-ex (Cass) that, and from your testimony here today, that insofar as the regulatory field is concerned you've had quite a bit of experience in the telecommunications field? A. Yes, telecommunications and electric power and the U.S. Postal Service and -- well... Q. Thank you. Your experience in the telecommunications field is both in the United States and here at the CRTC in Canada, I think we heard today? A. Yes. Q. Thank you. Would I be right in thinking that much of the analysis that we see in your evidence here in this proceeding is similar to analysis that you've applied in telecommunications proceedings? A. No. Let's distinguish the methods applied -- Q. That's what I meant when I said analysis, the methods. A. If you mean the methods applied, in general there is a substantial similarity. But, for example, in the telecommunications industry I didn't have occasion to consult partial measures of productivity, for example, such as the O&M-based measure, nor did I address anywhere in regulatory industries a case where output, for example, was measured by the number of customers and not by the quantity of services delivered to customers. Q. Okay. Fair enough. Thank you. So in bringing your analysis to the Canadian gas industry, did you direct your mind to differences in the Canadian gas Norsworthy 1024 cr-ex (Cass) industry that would be important for you to take into account in your analysis as opposed to other industries that you might have looked at in the past? A. Yes. However, (a), I'm not an expert nor do I claim to be on the Canadian gas industry, nor on the gas industry generally. I did, of course, consult with some of my sponsors in this proceeding, as well as some of my colleagues. For example, one of my students is just finishing a dissertation now on the local gas distribution industry in the U.S. and so we talked about some of the things that were (a), peculiar to the gas industry and (b), what might be different in Canada. Q. Okay. A. One of things that emerged in process was the pass-through in Canada that has not yet been achieved in the U.S. Q. Okay. So in bringing your mind to that analysis, what were the things that you determined to be important for purposes of productivity analysis that would be different insofar as the Canadian gas industry is concerned? A. Well, I think the most important difference is, as I indicated, the fact that gas costs are typically excluded and that the company has proposed a partial measure, a multi-factor productivity measure, based on much less than all of the purchased inputs and based on Norsworthy 1025 cr-ex (Cass) what I would view as an inappropriate measure of output. So these are some of the things that I considered. I guess there were others such as what the trends in gas prices had been in Canada compared to the U.S. and so forth, but that was just sort of sniffing the environment. It was very cursory. Q. I didn't intend to argue the company's proposal with you, Dr. Norsworthy. A. Oh, I'm sorry. Q. I just want you to explain to me that in doing your analysis you indicated to me that you tried to come to some understanding of the differences in the Canadian gas industry that would be important to your analysis. On a generic basis in relation to the industry, what are the differences that you saw to be important? A. Okay. Compared to industries generally, the capital intensity measured in terms of expenditures is much greater in the Canadian gas industry compared to, say, manufacturing industries or the postal service or airlines, the role of labour in total costs is considerably less. If gas is excluded, the role of purchased materials and services is considerably less than in other industries. By comparison with high-technology industries such as semi-conductors and computers and so forth, output by any measure grows considerably -- by any valid measure, Norsworthy 1026 cr-ex (Cass) grows considerably more slowly in the Canadian gas industry than it does in those other industries. I guess -- well, I looked at -- never mind. Those are some of the major factors that I observed. Q. Let me put a few to you. Would you agree with me that the rate of technological change and technological obsolescence is much greater in the telecommunications industry than the gas industry? A. Yes. Q. Thank you. Would you agree with me that there is a difference because of the potential for telecommunication carries to offer multiple services off the same infrastructure; whereas, the gas industry does not offer that same potential? A. Well, that depends just a little bit on what one means by different services. There are 10 or 11 different rate classes that are meant to reflect for Consumers Gas services that differ in some important dimensions that, therefore, warrant different unit costs -- I'm sorry, different rates based on imputations of unit costs. Q. I'm thinking about features that can be offered in the telecommunications industry like Call Waiting, Call Display, there is a wide variety of new services - I'm calling them new, I may be just outdated - in that industry, but we don't see that same phenomena in the gas industry, do we? A. No, that's correct. We don't see the same Norsworthy 1027 cr-ex (Cass) breadth of services. In terms of the number of services, it does depend a little bit on how you slice the chicken. Q. Now, in all your discussion of what you brought in your analysis in terms of a recognition of the specific features of the gas industry, I haven't yet heard you mention safety. Would you not agree with me, sir, that in a comparison between the telecommunications industry and the gas industry, the emphasis on safety is a significantly different factor? A. Certainly. Just as it is in the electric power industry. Q. Right. Thank you. A. I wouldn't want to say that, you know, that the electric power industry and the gas industry are strictly comparable in expenditures -- in safety expenditures per customer per unit of service, but safety is a much more important concept in gas and electric power than it is in telecommunications. Q. And I take it you would agree with me, sir, that - and I'm talking generically now of any industry - that the more a company succeeds in driving out the efficiencies that can be found in its business, the less potential it has to continue to drive out the same rate of efficiencies in the future? A. By "drive out", you mean take advantage of or, I guess when you say "driving out efficiency", you don't mean causing productivity to decline? Norsworthy 1028 cr-ex (Cass) Q. No, I don't. A. You mean it the other way? Q. Absolutely. A. Okay. Q. So you agree with my proposition on that basis? A. Yes, with a footnote. And the footnote is that to a very substantial degree efficiencies discovered are responsive to the incentives to discover efficiencies. In the telecommunications industry it's clear that there were very important incentives to reduce costs and if one listened only to what the company said one would have very substantially under estimated the performances that the companies were capable of and, indeed, they may have discovered that capability under the spur of incentive regulation. So my guess is that there are unexploited opportunities for innovation and efficiency improvement in the gas industry, as we are now observing in the electric power industry and so forth, and that incentive regulation will help to bring those forth. Q. Thank you. So I take it then you can agree with me that, everything else being equal, the company that has achieved a high productivity level, relatively speaking, cannot be expected to achieve the same level of productivity growth as those which are behind it in the productivity level? A. If the productivity level is measured Norsworthy 1029 cr-ex (Cass) appropriately, yes. Q. Yes. A. I haven't seen the analysis where the productivity level that Dr. Fuss refers to in his evidence, where that's measured. I understand by implication that it's measured in terms of customers as outputs and O&M inputs deflated by the Ontario CPI. So if that's the basis of the level comparison, I wouldn't be very comfortable with it. Q. All right. Well, again, I don't want to argue the case with you Dr. Norsworthy, but I think you've agreed with my proposition in a general sense, that the company which, relatively speaking, has achieved the higher level of productivity than its peers, everything else being equal, cannot be expected to meet the same -- or to maintain the same growth of productivity as those which are behind it; correct? A. If the level measure is appropriate, then yes, I agree with you. If we're speaking theoretically, I certainly agree with you. Q. Thank you. I wanted to ask you quickly some questions about the proposed formula. It's at Exhibit C, section 4, page 1. If you care to turn it up. They are such basic questions, I suspect you won't need it, but you might as well have it in front of you. A. Okay. Q. Now, you've given some evidence, Dr. Norsworthy 1030 cr-ex (Cass) Norsworthy, about labour outsourcing, so I want to talk to you about this formula in the context of labour outsourcing. A. Mm-hmm. Q. And through these questions I'm talking about the PBR period. So if we take the elements of the formula, for example, inflation: I take it you would agree with me that labour outsourcing during the PBR period is not going to affect inflation; correct, within the formula? A. The labour outsourcing wouldn't affect inflation? Q. Right. The inflation which is inserted into the formula. A. If inflation is in the formula taken as a proxy for the prices or the wages that the company pays for labour, combined with the prices of the materials input, then it would-- Q. Oh, sorry. A. --have that effect. Q. I probably misled you and made it far more complicated than it sounded. The inflation variable in this formula is proposed to be the Ontario CPI. A. Yes. Q. To the extent that the company does outsourcing of labour during the PBR period, the Ontario CPI is not going to change; correct? Norsworthy 1031 cr-ex (Cass) A. The Ontario CPI is not affected by the outsourcing of labour. Q. I didn't put the question very well the first time, I'm sorry. Same with the productivity that will get inserted into this formula during the PBR period; if the company does outsourcing of labour during the PBR period, the productivity figure will be set and will not change; correct? A. Not necessarily. It depends on what the prices are that are paid for the outsourced labour. O&M productivity is a ratio of outputs to inputs. The inputs, both labour and purchase materials, are deflated and the output is, in effect, deflated. So that it's sort of a measure of real output or physical output. So if we change the wage of labour, as might very well be the case in outsourcing, or if we change the quantity of labour services that are outsourced; that is, if a different quantity of services is provided by the unregulated affiliate, then, in fact, the MFP measure would be changed. Q. I'm sorry, Dr. Norsworthy, I may have misled you with my question again. I'm not talking about an estimation process, I'm talking about the operation of this formula in the PBR period. The Board will presumably determine an appropriate productivity figure and if there is Norsworthy 1032 cr-ex (Cass) outsourcing of labour during the PBR period, the productivity factor as determined by the Board will not be affected; correct? A. The target? Q. Right. A. The target will not be affected. Q. Right. And, similarly, during the PBR period customer growth is not going to be affected by outsourcing of labour; is it? A. It depends on whether -- do you mean the forecast of customer growth? Q. I mean the number used during the operation -- for the operation of the formula during the PBR period. A. If that is a forecast, that's correct; if it's an actual number, then changes in purchases, which may result in changes in the costs passed on to customers, could indeed affect customer growth. But, again, customer growth here is proxying for output, and so to that degree the formula is not really appropriate for a PBR plan. Q. Dr. Norsworthy, you've given some evidence on gas costs in your prefiled evidence and I gathered from your examination so far today that you have backed off, if I may say so, from your evidence in that regard. So if my next question is no longer relevant, just tell me so. But I want to take you to page 15 of your evidence, please. A. Mm-hmm, yes. Norsworthy 1033 cr-ex (Cass) Q. Looking at the second last line of the big paragraph starting at the top of the page, but I'm looking at the second last sentence of it. A. Uh-huh. Q. You say: 'In particular there may be arrangements whereby the company can obtain favourable long-term supply arrangements with suppliers.' I was going to ask you to elaborate on that with me. But is that no longer relevant now in view of the evidence you've given today? A. As far as I understand the company buys some gas for delivery to customers and there are intervening brokerage functions associated with other customers. To the extent that the company is buying gas for its customers, then this applies. But I am aware that there are other sellers of gas, other makers of market in the gas industry. Q. All right. Well then, please help me with what you would envisage the company doing then in the supply that its purchasing by way of entering into long-term supply arrangements. How long are you thinking of here? A. Well, let's look at it in another way -- Q. Well, is it difficult? Just answer my question. A. No, no. Q. When you say "long-term", what do you mean? Norsworthy 1034 cr-ex (Cass) A. It's very often the case that long-term arrangements of several years result in lower prices than, for example, buying gas in the spot market or now buying electricity in the spot market and so forth to sort of fill in the gaps. There are also options that can be purchased. Now, whether -- Q. I'm sorry Dr. Norsworthy, I thought my question was fairly simple. What do you mean by "long-term"? How long? A. Long enough to obtain favourable price treatment. Q. All right. So you wouldn't put any limitation on it then? The company should go out as long as would be necessary to get favourable price treatment. That's what you're saying? A. No. What I'm saying is that the company should seek to get favourable price treatment. The fact that brokerage is succeeding to some degree in the industry is a suggestion that there is either greater efficiency available in the brokerage function outside the company, or a different kind of contracting that may give these brokerage firms an edge. Q. Dr. Norsworthy-- A. Yes? Q. --with the greatest of respect, I still don't think you're answering my question. I've been asking you how long you think the Norsworthy 1035 cr-ex (Cass) company should go out on these long-term arrangements, and you gave an answer which was, in order to get the most favourable price on gas. A. No, I didn't say most favourable, I said favourable. Q. Well, a favourable price. And my proposition to you then was that what you're saying is there would be no limitation on the term, the company should just go out and -- A. No, I'm not saying there would be no limitation on the term at all. I believe that is something that you have inserted into the discourse. The fact of the matter is that there is no formal limitation on the terms of options in many commodities, but the market determines what the appropriate lengths of time are for those options or for contracts for the delivery of those commodities. And I would not attempt to prejudge that. I simply observe that the brokerage function has been successful in reducing costs in the electric power industry and the gas industry, that's all. Q. But I take it from what you have just said then that you would suggest that the company go out and tie up its gas supply for the long-term arrangements that would be dictated by the market forces that you've just described? A. No. That's incorrect. What I'm suggesting is that the company should seek more favourable arrangements. Norsworthy 1036 cr-ex (Cass) Q. And the way in which you have suggested that is by way of long-term arrangements; correct? A. Yes. That's one way. I perhaps should have said that there are other ways that it might do so, also. Q. All right. Let's leave it then. A. Sure. Q. Thank you, Dr. Norsworthy. Can I take you to Table 6 of your evidence at page 23, please? A. Sure. Q. Actually, I just want to be sure of the conclusions that you express blow this table. A. Mm-hmm. Q. You appear to be saying here that based on your presentation in Table 6, one can arrive at the conclusion that the industry and the company have had similar trends in output; correct? A. Yes, uh-huh. Q. And -- A. This is output measured in gas volume. Q. Yes. I appreciate that, yes. A. Yes. Q. And it's a revenue-weighted approach? A. Yes, uh-huh. Q. Thank you. And just below that, you also indicate that, as I understand what you're saying, that the company and the industry have also seen similar levels of decline in Norsworthy 1037 cr-ex (Cass) labour per unit of output; correct? A. Yes. Q. And, in fact, if we look at the more recent years of your data, speaking here of labour per unit of output - and I realize it's a little hard to get one's eyes on the correct columns - but if we look, say, for the period from 1991 to 1994 based on your data, labour per unit of output, the company has been doing better than the industry over that period, hasn't it? A. Yes, mm-hmm. Q. Thank you. So the nub of your evidence on productivity in relation to the company, Dr. Norsworthy, if it hasn't become apparent by now, is productivity in relation to capital; correct? A. I believe I say in at least one place that that is the most important factor in understanding the company's productivity decline. Q. Yes. Thank you. A. Mm-hmm. Q. And can you just help me with Footnote 12? It indicates that the data in Table 6 covers the period from 1987 to 1994. But then it indicates that you did an extrapolation for the industry; is that correct, that your industry data insofar as capital and labour inputs are concerned are actually an extrapolation beyond 1992? A. No. I think the term "extrapolation" there is probably inappropriate. In fact, the capital and Norsworthy 1038 cr-ex (Cass) labour inputs for the industry through '95 are those provided by Stat Canada, so that sentence is incorrect. Q. Well, the data doesn't go to 1995, Dr. Norsworthy. A. No. As I indicated before, because '95 is sort of a bad year for the company, I chose to terminate - that is, it was a substantial productivity decline that was made up later - I chose to terminate the comparison in 1994. Q. So just, you're clear in your mind then that there was no extrapolation because, just to be fair, I think we looked at the figures and it definitely appeared to us to be an extrapolation. But you're clear in your mind that that's not the case? A. I think if you look at the spreadsheet printout in page 2 of the supplement, you will see that the capital input and labour input are not extrapolations. Q. And that, you're clear about that in relation to this Table 6? A. Yes, mm-hmm. Q. Okay. Thank you. Now, Ms. Lea touched on this already, but given your emphasis on capital in terms of where you come to with respect to productivity, I'm just trying to grapple with what that means for the purposes of this case before the Board. Are you essentially telling the Board that it needs to go further in this case than the company's Norsworthy 1039 cr-ex (Cass) proposal? A. I'm sorry. Would you repeat the question, please? Q. Yes. Given your agreement, I believe, with what I had indicated earlier that the nub of where you come to on productivity for the company revolves around capital-- A. Mm-hmm. Q. --I'm trying to understand what your conclusions then mean for this case where there is not a mechanism in relation to capital before the Board. Are you telling the Board, are you suggesting to the Board that they should go further in this case? A. Well, I'm suggesting to the Board that the PBR plan should include capital as well as O&M inputs, yes. Q. So the answer to my question was a "yes"? A. Yes, mm-hmm. Q. Thank you. In relation to Table 6, we have heard some evidence about this perpetual inventory method and I just want to discuss that with you a little bit, if we can try to do it fairly briefly, Dr. Norsworthy, because you will appreciate I am a neophyte in this area. But as I understand the perpetual inventory method, in my terms what that would mean is that one would try to bring the capital numbers up to a market value or a replacement value. Is that fair? Norsworthy 1040 cr-ex (Cass) A. No. Actually, what it attempts to do is to express the whole series of capital in units that mean the same from year to year. In other words, it is a construction of a constant dollar or constant efficiency unit measure of the capital input. It does not seek to re-value it up to the present time. It seeks to value the whole stream of data in dollars of one particular period, so that they're comparable. Q. And you referred in your answer to using units that mean the same. Can we agree, Dr. Norsworthy, that in relation to your data for the industry presented in Table 6, that is done on a perpetual inventory method; correct? A. Stat Canada uses a perpetual inventory method to prepare that capital series, yes. Q. Right. And I think it's already clear from your evidence today the data for the company is not presented on a perpetual inventory method; right? A. Absolutely, yes. Q. And you indicated that Statistics Canada uses the perpetual inventory method; correct? A. Sure. Q. I believe the U.S. Bureau of Labor Statistics uses that method; correct? A. No. Actually, it doesn't. It uses a variant that includes some sort of strange treatments of Norsworthy 1041 cr-ex (Cass) depreciation. Q. All right. But it uses a variant of that method? A. You could say that. Q. All right. Thank you. And the CRTC has been using that method; correct? A. Sure. Q. Thank you. THE PRESIDING MEMBER: It's twelve o'clock. I think we should stop. Do you know how much longer you would likely be, Mr. Cass? I don't want to... MR. CASS: I am coming towards the end, Madam Chair. I would say on the outside I would hope to keep it under half an hour. THE PRESIDING MEMBER: Okay. We will try and resume promptly at two o'clock and I think we will investigate whether we may have to sit a little bit late this evening. MR. CASS: Thank you, Madam Chair. MR. JANIGAN: Madam Chair, before we adjourn for lunch, I would like to convey my apologies for our absence after the break. We grossly underestimated the time that it would take to complete some business with another part of the Board and I apologize for the delay. THE PRESIDING MEMBER: Thank you, Mr. Janigan. So we will reconvene at two. ---Luncheon recess at 12:00 p.m. Norsworthy 1042 cr-ex (Cass) ---On commencing at 2:00 p.m. THE PRESIDING MEMBER: Please be seated. Any preliminary matters? ----(No response) Mr. Cass? MR. CASS: Thank you Madam Chair. In the time over lunch I realized that at the rate we were going with the ground I had yet to cover that my time estimate was in jeopardy and maybe the afternoon was in jeopardy, so I've tried to cut it down to a bare minimum of questions and see if we can move through it quickly. MR. CASS: Q. Dr. Norsworthy, in some of your evidence this morning, I think perhaps more than once, you've indicated, in Table 6, for example, you left out 1995 because that was -- I'm not sure how you characterized it, a bad year for the company; is that correct? A. Yes. Q. And you'd be aware that that was a year of very warm weather? A. It was in northern New York. Q. So you have nothing that would cause you to disagree with the notion that warm weather was an important factor in that year? A. That's right. Q. Right. And so in your approach where you are using volumes as the output measure, I take it then that Norsworthy 1043 cr-ex (Cass) you're agreeing that warm weather can have the effect of distorting the productivity level, correct? A. In the short-run. That's why generally for establishing incentive regulation plans and whatnot it makes sense to look at a longer period of time. The way that we do the calculations, it is especially sensitive to the terminal year and the initial year, and so we try to be careful about that. Q. So if, for example, in the short-term, as you said, if the company had a year where because of weather its volumes were down by 10 per cent, one would expect on your approach a relatively proportionate reduction in productivity, correct? A. Certainly -- yes, if you say relatively proportional I can buy that. Q. To the extent that one looks at the longer term, if in fact the company has been experiencing warmer weather over the longer term, then in your approach that's going to flow through to the productivity performance, isn't it? A. If there is any kind of a secular trend toward warmer weather one would expect that the company would adjust its inputs so that it's productivity decline would be mitigated. Q. Okay. I'll just keep moving right on. I just want to ask one question about the materials deflator, Dr. Norsworthy. You'd agree with me in your evidence with respect Norsworthy 1044 cr-ex (Cass) to the appropriate materials deflator, you've not endeavoured to provide any comparison to satisfy us, to satisfy the Board, that the underlying mix of goods and services in the deflator that you propose corresponds with those in the company's estimate of materials, correct? A. I'm not sure you'd say that I've made no effort because we did request detailed materials expenditures from the company. Q. But, in any event, you haven't put forward that evidence that I've referred to? A. Absolutely not because, well, we didn't get that information. Q. Okay. And just very quickly then, Dr. Norsworthy, I want to come back to our discussion of your analysis of the company's capital productivity as compared to the industry's capital productivity and hopefully run through quickly some questions. I just want to confirm whether or not you adjusted for certain factors. We don't need to argue about them now, but I just want to know whether you have adjusted for them. I take it from your evidence already that in this work that you did with respect to capital productivity there was no adjustment for conservation, correct? A. That's correct. Q. Right. I take it as well that you have not adjusted in any way for the Enbridge Consumers Gas cast iron main replacement program, correct? Norsworthy 1045 cr-ex (Cass) A. That's correct. These are properly -- in the normal measurement of productivity, these are factors that would be reflected in the productivity data and not adjusted out. Q. Again, we can argue later. I just want to find out -- A. I think being clear, to be perfectly honest, is not argument. I'm simply trying to clarify the answers to the question that you ask. Q. Okay. And you have not adjusted in your capital productivity analysis for the high growth rate in Consumers Gas franchise areas, a relatively high growth rate, particularly in respect of residential customers, correct? A. Well, I think that growth is reflected in the measure, certainly. Q. But you have not made any adjustment for it? A. Agreed. It is also the case that in normal productivity analysis no adjustment is called for. Q. And you've not made any adjustment for normalized weather, correct? A. That's correct. I believe I have said before that such an adjustment would not be made in the usual case. Q. And you've not made any adjustment for the impact of conversions of gas customers from company service -- from sales service to T-service, correct? A. That's absolutely correct. Norsworthy 1046 cr-ex (Cass) Q. Okay. And now I just want to quickly take you to, still in this same line of questioning, to the two references that were provided to your counsel yesterday. One is an excerpt from a book entitled "The Politics of Energy" by Bruce Dern and Glen Toner, pages 371 through to-- A. 397. Q. --397? A. Yes. Q. If you've had an opportunity to read that you would have seen at page 387 under the heading "Natural Gas Laterals Program and the Gas Marketing Assistance Program", the first sentence there, which is a discussion of very large assistance provided in the Province of Quebec for lateral construction. Did you have a chance to see that? A. No, not really. I think in order to read this properly I would have to -- well, it's almost like reading a U.S. Department of Defence document. We've PIPS and CHIPS and COSPS and HEMS and so forth. So I would say that to comment intelligently on this I really would have spend a little more time on it. Q. I appreciate that, and I'm not going to ask you to comment, Dr. Norsworthy. I am on my same line of questioning about adjustments, and we see there, just adding up the numbers roughly, probably something in excess of $400-million in assisted capital in Quebec just prior to your study Norsworthy 1047 cr-ex (Cass) period. And my question simply is to confirm that you made no adjustment for that in your calculations? A. No. I'm not sure why -- does Consumers Gas also serve Quebec? Q. Well, Dr. Norsworthy, again, I'm trying to keep this moving fairly quickly -- A. So am I. I'm trying to understand why you asked the question about Quebec. Q. Well... A. Okay. Q. You can agree with me you didn't make an adjustment? A. Absolutely. Q. All right. Thank you. Another document was provided to you with respect to the Federal Government Infrastructure Program which has been in place in the 1990's in Canada and the document revealed spending on gas infrastructure in provinces other than Ontario, but no such assistance in Ontario. Did you see that document? A. I see the document. Q. Again, no adjustment was made in your calculations for that? A. No. I would have to say that there is no evidence that comes to my attention that such adjustments would necessarily be appropriate or even optionally be appropriate. Norsworthy 1048 cr-ex (Cass) MR. CASS: That completes my cross-examination, Madam Chair. Thank you. THE PRESIDING MEMBER: Thank you, Mr. Cass. The Board has some questions. Dr. Higgin. DR. HIGGIN: Thank you. EXAMINATION BY DR. HIGGIN: Q. Good afternoon, sir. A. Good afternoon. Q. Could I ask you to look up a piece of Dr. Fuss' evidence and it's at Exhibit C, Appendix B, and it is page 8 of 8. If somebody could assist you with that. Just while you are finding it, this is a mathematical derivation of the formula based on what I believe are some accepted indices that have been developed by theory, economic theory for productivity. I'm going to focus on the productivity area. So if you have that, could I ask to look at equation A-6 which defines productivity. A. Yes. Q. Just by way of background, it was clarified with the company's witnesses that that was the formula that they utilized in order to calculate the productivity -- historic productivity of the company. A. Well, I presume a slightly different form must have been used because there is an unmatched left parenthesis in this expression. Q. Yes. I'm assuming that there is a missing -- I assume, maybe naively, there was equally as well a Norsworthy 1049 ex (Higgin) missing right parenthesis. A. That's another way to think of it. The left parenthesis I think can be eliminated without loss of honour. Q. Well, I thought you may be familiar with the authors that lay behind this, the so-called Tornqvist index. Are you familiar? A. Yes, I am. Q. My question relates to what would be the additional factors that you would recommend go into this if you were going to do it for total factor productivity, as you have indicated your preference; and secondly, there is, of course, also the fundamental difference of what "y" is, the output-- A. Yes. Q. --the output factor. We'll talk about "y" in a minute, but if you just could give me a rough indication of what type of other parameters you would recommend would be in the formula for productivity, in a total factor productivity that you feel is appropriate to this firm, this company? A. Yes. A total factor productivity measure that added capital in a way that's completely parallel here to the treatment of labour and materials would be such a measure. Now, I would caveat that by saying that my productivity calculations are based on a Fisher index rather than Tornqvist index, but there is very little Norsworthy 1050 ex (Higgin) difference at this level. The difference really comes when new services are added or old services disappear. Q. Would it be similarly the change in the share-- A. Yes. Q. --of -- A. Well, it would be the change in the log-- Q. --from year to year? The log? A. --weighted by the average share-- Q. Yes. A. --over a two-year period, yes. Uh,huh. Q. Now, just turning briefly then to the productivity factor, you have disagreed with, of course, the use of customers as an indicator of output? A. Yes. Q. And I'm just trying to get a feel from you about the differences between a utility that has a, say, heavy number -- small number of customers -- same volume, let's assume the same volume? A. All right. Q. Just a hypothetical. One that has a very large number of -- has a number of large industrial customers making up that volume as opposed to, in this case, over a million small volume customers? A. Uh-huh. Q. Isn't there some convergence in that case between the use of volumes and customers as output? Norsworthy 1051 ex (Higgin) Don't they get closer together in a situation where there is a large number of small volume customers? A. They do get closer together, but the appropriate measure from the perspective of productivity practice is still to use the revenue weighted volumes as the measure of output; that is, insofar as the small customers don't change their use of gas, then the two would be very close together. But, as counsel for Enbridge pointed out, if gas per customer is declining, then there will be a decline in productivity measured by volumes; and if gas per customer is rising, then the reverse would happen. DR. HIGGIN: Thank you very much. Those are my questions. Thank you, sir. THE PRESIDING MEMBER: Mr. Vlahos. MR. VLAHOS: Doctor, just a question or two. EXAMINATION BY MR. VLAHOS: Q. I wasn't sure how to -- if you were to measure the output on a volume measure-- A. Mm-hmm. Q. --and you say it has to be weighted by your revenue. Can you just help me visualize this-- A. Sure. Q. --as to how you would do it. And you may want to take an example -- an easy example of two rate classes, one customer versus 10 customers, and give us an idea as to how that would be calculated. Norsworthy 1052 ex (Vlahos) A. Sure, okay. The principle of weighting by -- weighting volumes of service delivered by unit revenues is quite common, it's not -- you know, we're not innovating here in suggesting the use of that approach. Rather, the innovation is from the other perspective, the customers might be appropriate. Now, suppose we have a volume of 10 units going to one class of customer and a volume of 20 units going to a second class of customer, and let's suppose that the 10 units are delivered under circumstances that require more distribution plant in place. Then, in general, the rate-making process will recognize this and authorize higher rates for the class of customers consuming only 10 units, higher rates per unit of gas. And so, correspondingly, the relative importance of that 10 units of gas in the ultimate output measure will be greater per unit than the 20 units of gas which are delivered under lower cost consequences. Now, this assumes in each case that something like the cost of service is used to determine rates for both classes of service. So, under those circumstances, when there is more -- when one of these classes of service, the volumes delivered under one of these classes of service increases and the other stays constant, there will be a change in the mix of gas volumes and there will be a change in the Norsworthy 1053 ex (Vlahos) average revenue per unit of gas, and what we will get is more or less output according to whether the growth took place in the higher cost or the lower cost category of service. So, once again, if the high cost category of service rises relative to the low cost category of service, we're going to wind up with proportionately more output than just the percentage increase that that category of service would suggest. Q. Okay. And is a starting point though, if I have -- let me just give you some easy numbers. If I have two classes, a Rate 01 and a Rate 2. A. Mm-hmm. Q. And I have volumes of 10 units for Rate 01 and say 20 units for Rate 2. And let's say revenues for Rate 01 will be $15. A. Mm-hmm. Q. And for Rate 2, I guess it has to be something relatively smaller, so let's make it $21. A. That is total revenues? Q. Total revenues, yes. A. Okay. Q. So you get the unit rate and it would be -- in the case of Rate 01 you would have $1.50, and in the case of Rate 2 you would have a-dollar-something. A. A dollar five. Mm-hmm. Q. Okay? What is my base now? Can you help me understand Norsworthy 1054 ex (Vlahos) my starting calculation before I go to the changes? A. Okay. In this case we have total revenues of 36 units, and so the initial output would be 15 over 36 times 10 plus 21 over 36 times 20, and that is your base output index. Now, I don't know if you want me to...? Q. No, that's fine. A. Okay. Q. That's fine. Just my last question. From the analysis that you have done and your recommendations and your suggestions, is any of the indices or the data that would be required for the Board to implement or to accept the PBR mechanism, any of those indices that you have worked with that are not available as readily as CPI? Does someone have to do some more work than simply go to StatsCan publications or go to some kind of service and just pull that data right away? A. Well, Stat Canada is in the process of revising, as I indicated, it takes a little bit longer to get the data. Also you can't simply open a page at this time, you have to request that the data be sent and, in order to do that, you have to find out exactly what data you want. But, yes, the data are publicly available to undertake the analysis. Q. I just want to get a feel as to how much work would be involved in comparing have the CPI applied to the Norsworthy 1055 ex (Vlahos) formula as opposed to some other index, as you would suggest. A. It takes more effort to calculate the materials price index based on data from Stat Canada than it does to simply open a book and reference the CPI; however, the difference, as I indicated, over a period of several years turns out to be $6-million which may be an important number. Q. And that is data that someone presumably can dig into and come up with. Is that easily verifiable? Can that be easily tested in a hearing room like this? A. I think so. I think that what you might want to do is perhaps have Rene Durant from the Productivity Division validate that the measures are appropriately derived and applied. But that would be something to which you could have some such person do a quick review and make sure that no egregious errors had been made. The calculations themselves are straightforward and captured in a spreadsheet. Q. Okay, Doctor. And, in the same vein, if we turn to the growth factor-- A. Mm-hmm. Q. --compared to what the company proposes, what are the additional implications of having your growth factor estimate rather than the company's? A. The output growth factor and the input growth factor and so forth? Norsworthy 1056 ex (Vlahos) Q. I'm sorry, the customer versus -- A. Ah. Q. -- versus the volume. A. That's not necessarily an easy question. A fairly easy one is, you know: What are the differences that you get? And I've documented that in my testimony. Ultimately, the Board has to decide what sort of target it wishes to incorporate in the PBR plan. And as I think we discussed earlier in responding to questions from the Board Staff, there is no substantial savings in regulatory burden and relatively minor savings in calculation provided the company makes available data that indeed it must in order to file its taxes -- that is, data that it must have in order to file its taxes. Under those circumstances, the calculation burden is pretty straightforward, it's well defined, and the regulatory burden is less if the measure includes total factor productivity; that is, the capital as well as the labour and materials of O&M. Q. But sticking with the targeted PBR as the company proposes. A. Okay. Q. If I had to replace the customer measure-- A. Mm-hmm. Q. --with the measure of volume-- A. Hm-hmm. Q. --weighted by revenues the way you have described-- Norsworthy 1057 ex (Vlahos) A. Yes? Q. --I'm just trying to get my mind around to what are the additional areas of testing that would have to take place and whether the benefits that may come out would justify the additional effort or cost? A. Okay. Well, I calculated the volume-based -- the revenue-weighted volume measure of output, and it was not terribly difficult once the data were in hand. And, in fact, the data were supplied by the company. So I don't see that there would be an additional calculational burden. The resulting productivity measure initially would show a lower result, both because the output would grow more slowly and the appropriately deflated materials input would grow somewhat more rapidly. It is my speculation - and I'm quick to label it as that - that in the relatively near future the productivity will begin to rise, both as a result of what will hopefully be a properly designed and applied PBR plan, and also as the customers sort of catch up, the customers' usage catches up with some of the recent capital expenditures. Q. Okay. Just to follow up on that then, the productivity measure will initially be lower and, everything else being equal, that means that the base O&M would have to be higher? A. The measured O&M input, yes, would be somewhat higher, yes. Norsworthy 1058 ex (Vlahos) Q. And if there's no way or avenue to change the productivity measure - and you may be aware, the company's proposing that the productivity measure that it has recommended should survive for the duration of the PBR term - would that give you a concern, that we have lowered the productivity measure, we have increased the O&M base, and there's no recourse for three years? A. Well, I think if you basically make that change in the way output is measured and in the way materials input is measured, then from that standpoint, you would select the PBR target. In other words, you would take into account the fact that O&M productivity measured in this different way should be accommodated by a different productivity target with whatever stretch factor is appropriate and so forth. Q. All right. But do we say that it will be initially lower? A. Yes. That's right. Q. Okay. Would the inflation factor -- it will change? It will not be CPI, according to your suggestion; it will be something different? A. Yes. That's right. Q. And that something different will be a lower...? A. A lower rate for price inflation for the inputs, yes. Q. Is there a complete outset between the initially lower productivity factor and initially lower Norsworthy 1059 ex (Vlahos) inflation factor or is there a difference there that has to be made up from the left-hand side of the equation? A. If you are looking at a constant "X" factor, then reducing the materials input price triggers an offset in the input price differential that exactly balances. However, from the perspective of appraising the productivity performance of the company, you want to look at the productivity growth rate separately, so that when you change the output measure and change the input measure, then you have a different historic pattern of productivity that you would use to infer what you want the "X" factor to be. MR. VLAHOS: Okay. Thank you for those answers. Thank you, Madam Chair. THE PRESIDING MEMBER: Thank you. THE WITNESS: Thank you. THE PRESIDING MEMBER: Just one question which is a little bit a more general comment on the kinds of things Mr. Vlahos was asking you about. EXAMINATION BY THE PRESIDING MEMBER: Q. I didn't understand, from anything you said either to Board Staff counsel or in response to Mr. Vlahos' questions or in your evidence that you thought you could just change the indices, use the targeted PBR, and come to some kind of reasonably right answer; right? Although, if you're going to use the targeted PBR and you do recommend different indices, I didn't see anything in your evidence which suggested that you didn't Norsworthy 1060 ex (Presiding Member) stick to your guns really and say the capital has to be in there, as well. A. No. That's correct. THE PRESIDING MEMBER: Okay. Thank you. Mr. Janigan? MR. JANIGAN: I have no re-examination, Madam Chair. THE PRESIDING MEMBER: Thank you very much. Thank you for your assistance. THE WITNESS: Thank you very much. ---Witness withdraws MR. WARREN: Can I ask Dr. Bauer to come to the stand, please? If I could ask Dr. Bauer to be sworn. I don't believe he has been. JOHANNES M. BAUER, Ph.D.; Sworn. MR. WARREN: Madam Chair, Mr. Cass has graciously agreed that I can forego the ritual of Dr. Bauer's qualifications, which I appreciate. And with your permission, I wonder if I could just enter him as a witness who is an expert in matters of PBR for regulated utilities? THE PRESIDING MEMBER: Yes, Mr. Warren. That will be satisfactory. DIRECT EXAMINATION BY MR. WARREN: Q. Dr. Bauer, there is certain prefiled evidence in this case which has been marked as Exhibit 2.1. Was that exhibit prepared -- was that prepared by you? Bauer 1061 dr ex (Warren) A. Yes. Q. And are there any changes you wish to make in it? A. Yes, there are a few editorial changes. On page 6 at the end of the top paragraph, the fourth line, it should read "remainder of this evidence". There's an "a" missing. On page 8, in the first long paragraph, the sixth line from the bottom of this paragraph, there's a colon and it says, 'The growth factor will have to represent...' and then it should read "the" instead of "to capital costs". Then on page 9, the first full paragraph on this page, the fourth line from the bottom of the paragraph, it reads, "namely the traditional PBR". That should read "namely the traditional ROR". It's the traps that you run into if you use similar acronyms. MS. LEA: Where? THE WITNESS: A couple of them are tongue-breakers. Let me see. Then on page 12 in the paragraph that is flowing over from the previous page, the third line from the bottom of this paragraph, I refer to the Canadian CPI and then in parentheses, I say Z-CPI, and that is inconsistent with the graph that reads D-CCPI for different Canadian consumer price index, so this should be corrected. Bauer 1062 dr ex (Warren) And last, but not least, in my recommendation No. 3, on the first line, it should read -- MS. LEA: Sorry. Page, sir? THE WITNESS: Recommendation No. 3. That's page 21. MS. LEA: Thank you. THE WITNESS: It says, 'If regardless the historical company performance is selected...' There's an "e" missing. These are all my corrections. MR. WARREN: Q. Dr. Bauer, I believe you were also responsible for certain interrogatory responses. Were those responses prepared by you and do you adopt them in this case? A. There was one addition in one of the interrogatories, and I have to admit I don't have the number with me. I was asked about whether I was aware of any studies looking at productivity developments in the gas industry in Canada. And at the point I answered the interrogatory, there was no such study. In the meantime, there was Dr. Norsworthy's piece. Q. Do you adopt both your prefiled evidence-- A. Yes. Q. --and the interrogatory responses? A. Yes. Q. Briefly, some points I wanted to cover arising out of the testimony that we've heard over the Bauer 1063 dr ex (Warren) past 10 days or so. First of all, Dr. Bauer, Dr. Fuss, when he was here the other day, spoke about what's known as the Averch Johnson theory and he, albeit quite gently, attempted to put it to rest and left us with the impression that one should visit it in must the same way one visit Lenin's tomb; that it's now embalmed and of not much value. Could you advise the Board whether or not you are, (a), familiar with this and whether, in your view, it has any currency? A. Certainly. The original piece by Messrs. Averch and Johnson was written in a particular context of full rate of return regulation in the 1960s, and as we gradually move away from that environment, some of the issues that they raised also disappear. So therefore, it's not surprisingly that there has been less attention paid to their original paper; however, the issue as such has not disappeared and it is discussed in a number of contexts. You know, just a recent search in the top academic journals reveals that at least, well, 20-plus papers have been published in the last five or six years that cover issues like this, and these are highly prestigious journals. Interestingly, there was one published just a few months ago that looks at the impact of regulation on input price substitution in the natural gas industry in the Bauer 1064 dr ex (Warren) Southern Economic Journal that comes to the conclusion that the input distortion actually led to a 16 per cent increase in the costs over efficient levels. So to say that the discussion is essentially dead I think is somewhat wishful thinking. It's probably not the height of the debate any more, but there's continuing work in this area. There's a second aspect of the paper that commonly was ignored in the past, but receives more attention is there is notice that in the paper, it was argued that utilities that work, on the one hand, under regulated conditions and on the other hand in more competitive conditions, may actually have an over-incentive to diversify. And that is of renewed interest. It was irrelevant essentially in the 1960s because most utilities were under comprehensive rate of return regulation. Given now our diversification situation where we have utilities more and more operating under these conditions, this aspect of the paper has been investigated many, many times. And so I think it is inaccurate to state that the issue is dead. The original question has become largely irrelevant, but the issue is not dead. Q. This morning during Mr. Cass' cross- examination of Dr. Norsworthy, he pointed him to an exhibit which has been filed in an interrogatory response, and you don't need to turn it up. It's Exhibit D, section Bauer 1065 dr ex (Warren) 4.44, and it is a paper prepared by Drs. Lowry and Kaufmann entitled "Forecasting the Productivity Growth of Natural Gas Distributors". Are you familiar with that paper? A. Yes, I am. Q. And can you tell the Board whether or not that paper has been considered by regulators in the United States, and if so, in what way? A. Yes. It has been considered in a number of cases. I'm aware of, for example, the Boston Gas case in Massachusetts. And the paper in that proceeding was essentially criticized for using the wrong output measure. It is using the number of customers as the measure of output and it was, for that reason, essentially discredited or at least, you know, it didn't receive full acknowledgement in the proceeding. And the Massachusetts Department of Public Utilities modified in the end its productivity significantly with two factors. One was a consumer dividend of 1 percentage point and, in addition, they added a 1 percentage point accumulated inefficiencies, as they call it, factor to take care of inefficiencies that they assumed were present in the past, although they couldn't really pinpoint those. Q. The next point I wanted to address is that during the course of the testimony in the past week or so, Enbridge Consumers Gas has articulated its objection to a Bauer 1066 dr ex (Warren) comprehensive PBR based principally, though not solely, on the alleged difficulty in accommodating capital investments which do not recover costs quickly. And in doing so, they pointed to what they identified as a public policy component in that some investments are made pursuant to an obligation to extend mains and to ensure safety which are not economic in the short run. I'd like, if you wish -- if you can, Dr. Bauer, your views as to whether that concern is one which can be overcome in the design of a comprehensive PRB? A. I don't see any conceptual difficulty in overcoming this concern. There are a number of practical examples in existence where capital has been followed into the PRB plan, again the Boston Gas case comes to mind, the Southern California Gas case comes to mind and the B.C. Gas example comes to mind, and although there are complicated issues to solve, I think they are not insurmountable obstacles. Actually, anecdotal evidence has it, that I'm aware of, that in the B.C. Gas case the stakeholders were able to come up with the capital part of the proposal in about five days. Q. Now, I want to turn to the issue of rebasing. MR. WARREN: And, Madam Chair, Dr. Bauer has illustrated his comments in this area and in some succeeding areas. He has illustrated by some printouts which I'd ask... Bauer 1067 dr ex (Warren) These materials, Madam Chair, are all produced data which is on the public record in this case already. Q. As a prelude to these questions, Dr. Bauer, there has been much discussion during this case about the benefits which flow to ratepayers and the focus for some, if not all, of that discussion has been the issue of productivity gains. The company is proposing a productivity factor of 0.63 per cent and there has been some testimony given to the effect that it may be able to achieve productivity higher than that to the question of why ratepayers should not get some portion of the benefit of the productivity above 0.63 per cent. Consumers responds that ratepayers will get the benefit of the actual productivity when the O&M budget is rebased at the end of the third year of the PRB period. Can you, Dr. Bauer, provide -- do you have any comment on whether or not, in your view, consumers are assured of the benefits from the rebasing that is referred to by Consumers? A. In the current proposals consumers are not assured that they will benefit from any rebasing simply because we don't know at this point how this rebasing will take place. Q. Dr. Bauer, when you refer to "consumers", do you mean ratepayers? A. Ratepayers, yes. And the issue is actually more complicated and, Bauer 1068 dr ex (Warren) if I may, let me just use this one diagram, the first diagram, that I prepared just to highlight or to frame our debate, to put some flesh and some meat to our argument. What I did here -- MR. WARREN: Madam Chair, can we have this marked as an exhibit so that the record makes sense when we ultimately read it. MS. LEA: Thank you. Exhibit I7.1, and I understand Dr. Bauer is about to refer to the first page of Exhibit I7.1. THE WITNESS: It looks like little a stepladder here. MR. WARREN: It is entitled "Benefits of -- THE WITNESS: Yes, it is entitled "Benefits of Indexing". ---EXHIBIT NO. I7.1: Document entitled "Benefits of Indexing". THE WITNESS: What I tried do here is simply illustrate for us so that we can think through this more clearly how a scenario could like and under what conditions rebasing would be feasible and what the implications of rebasing would be. What I did here was I used the O&M base that was submitted in Consumers' prefiled evidence that has shifted over time, and I have to admit I'm not clear if this is still the value that we are working from. What I did then, just to make those assumptions clear, is I assumed following the following things. Bauer 1069 dr ex (Warren) First of all, I assumed that we would use the O&M PBR formula as proposed by the company. So it adjusts this initial base of customer growth for productivity growth, as well as for inflation, measured based on the company's proposal. Secondly, I assumed just for sake of simplicity that there is an inflation, an expected inflation, the Ontario CPI of 2 per cent per year. I'm trying to simplify here. What the thick line shows you then is what the adjusted O&M budget would be that is derived from using the formula. So it would increase to 243.9 in year one, 247.2 in the follow year and then 250.5 in the following year, because I assume we know the inflation data for sure in consumer growth. In this case I have set it at zero just to make things easier. There is no rebasing at the end of the year that would have to take place to take account of deviations between the forecast inflation rate and the actual inflation rate. This really is sort of simplifying things to the level to make it transparent. Now, the big question is, of course: What is the company's actual or feasible productivity growth in this time? And, again, to make things easier I have assumed that the company can achieve what it has indicated in one of the discussions here last week that there is approximately 2 per cent per year productivity growth. If that is the case, and the inflation rate that I assume in this example is also 2 per cent, it means that Bauer 1070 dr ex (Warren) the O&M expenditures of the company actually remain constant throughout this three-year period. That's indicated in this dotted line that you see, the horizontal line. Now, let's think briefly now what the benefits would be to the shareholder in this scenario, for one. Obviously, in the first year there would be benefits of 3.3-million; the difference between the test year O&M derived from the formula and the company's actual O&M expenditures. In the second year, I look at the difference of 6.6-million and the third year of 9.9-million. Okay? Now, these benefits flow to the shareholder. Now, the big question is what will happen at the end of the three years? Will Consumers reveal that its true O&M expenditures are still only 240.6-million? When it is argued that rebasing will take place, we implicitly assume that at the end of these three years we will know what the O&M figure is, and this is not a trivial assumption because, according to the current plan, I don't see any reporting requirements for this three-year period. At the end of three years, the Board may well be in a position where it cannot meaningfully assess what the true O&M cost structure of the company is. So information deficit may be problem. Secondly, the company at the end of this period could come in and say: Well, this was only reflective of Bauer 1071 dr ex (Warren) past events. In the future, our O&M expenditures will be higher for whatever reasons. And I'm not making this up as a theoretical argument. There are cases where we know -- we went through a rebasing of the PBR plan where this argument routinely was brought forward. Thirdly, the company could say: Well, the best strategy from now on would be simply to continue the plan. Again, in the case in Canada, West Kootenay Power, for example, that's exactly how they argued. They said just leave it untouched, let's continue it for another three years. So we cannot at this point assume that the rebasing, if even feasible, if we could overcome this information deficit would take place. The only way we are guaranteed such rebasing is at this point in time when the plan is implemented there is a commitment that this rebasing will take place according to a specific process. I think if the process is not defined at this point in time today or when this decision is made, then we cannot really argue forcefully that such rebasing will take place to the benefit of consumers. Q. Can you, just as a corollary to that, Dr. Bauer, suggest what conditions the Board might set down with respect to rebasing that would cover off those concerns? At least broadly sketch them. A. Well, for one, I think one would have to assure that operational and cost data are being reported Bauer 1072 dr ex (Warren) to the Board in the period of plan so there is not a three-year gap in information that the Board has available. Secondly, there needs to be a commitment to indeed rebase to the true O&M level, and I think that the traditional rate case format would probably be an appropriate forum to conduct such a case, but let me emphasize here that there are many people testifying in regulatory proceedings that say it is the inappropriate way to review a plan after three years. So, in other words, in those three years, if there is no commitment today, there could be a shifting of perspective and people could come in here and argue that using the rate of return methodology to reveal true O&M expenses in three years is inappropriate because it is incompatible with incentive regulation. There has been a lot of debate on the problem that the regulator may unilaterally reset the parameters of an incentive plan that has been discussed under the heading of "Regulatory Recontracting", but the symmetrical problem really exists from the corporate side, too. In this case here, where there is a promise that in three years we will rebase, we need to have more, I think more support in the form of a commitment to place trust in it. Now, having said all these things, there is merit in the argument that if the incentive plan indeed incents the company to reveal efficiencies that they would not Bauer 1073 dr ex (Warren) show on the cost of service regulation and that we would know at the end of the three-year period what the increased efficiencies are, the ratepayers will be better off at the end of the period. But this, as I said before, this is based on the very strong assumption that we can indeed rebase and will rebase in the way this is seen. Regardless, though, the fact is that if you look at the magnitude of benefits that will be derived from the simple scenario, and you can play around and vary the assumptions and you will see that one fact will always hold, that the sum of the benefits to the shareholders in these three years is always, is always larger than the rebasing effect at the end of the three years. Now, you can say it's the cost, that's my investment cost of learning what the increased efficiencies could be. But I want to highlight this: There is only one scenario when the benefits to the ratepayer are higher than the costs to the shareholder and that is if the company every year does worse than this indexing formula would predict. Q. The next topic I wanted to move to, Dr. Bauer, is there has been discussion in questions that were posed by, among others, by me and also by Dr. Higgin at one stage, asking for Enbridge Consumers Gas witnesses to compare the targeted PBR proposal with the existing cost of service regime. Bauer 1074 dr ex (Warren) I wonder, Dr. Bauer, if you could comment -- and I should say that the discussion has taken place in the context of, among other matters, issues of reduced regulatory burden, the benefits of productivity and the rebasing matter that you've just talked about. I wonder, Dr. Bauer, if you could comment in your view whether the -- as the plan is before the Board, in your view, ratepayers will be better off under this proposal or under the existing cost of service regime? A. There are several areas that one has to look into. One relates to, let's say, the administrative cost of regulation and there has been, repeatedly there have been claims that PBR would reduce the cost of regulation, and I think it has been said earlier in this proceeding by some of the witnesses that typically what we see, especially in narrow-based PBR, that new issues arise, complicated issues as to the link, for example, between the capital expenditure side of the company and the O&M side that, therefore, the record, as we know it, shows that the cost of regulation that they spent negotiating things actually does not go down, and so that was probably a premature hope that many people have. Most people who have studied empirically what happens on a PBR have come to the conclusion that the costs of regulation actually do not significantly decline as long as we are in such hybrid scenarios. In other words, still we use a lot of the rate of return tools and only cautiously PBR tools. Bauer 1075 dr ex (Warren) I think the conclusion does probably not hold if one would move forcefully to a more comprehensive plan, especially of the price cap type which is, you know, as I have said many times before, my clear favorite in terms of regulatory mechanisms. The issues that would have to be addressed, of course, in regulatory proceedings are things such as were they indeed legitimate concerns that the company substitutes capital expenses for O&M expenses. I don't think that this is a moot issue. I think the way how, as far as I know, the information technology has been treated recently where initially it was planned as an O&M component now shifted to capital should raise flags I think of caution. Secondly, the "Z" factor adjustment. In the current proposal I think there has been an open field to bring whatever is seen as inconvenient up as a "Z" factor adjustment which could lead to prolonged debates on these things. And last, but not least, the service quality issue which again, in the current proposal, I think there is a lot of uncertainty on how service quality issues will be dealt with once there is a diagnosis that the target has not been met. So I'm at this point very skeptical whether this promised reduction in the regulatory burden will happen. Secondly, I think it's worth asking the question whether the marginal differences that we see between the Bauer 1076 dr ex (Warren) targeted PBR plan and cost of service regulation really warrant some of the learning that is necessary in implementing a new approach. And I did a -- really just sort of an historical experiment, if you wish, and I said: Well, let's just try what would have happened in the past 10 years if we applied the proposed formula to the O&M budget starting in 1987 and see how it would have evolved over time. And this is what you see on this diagram, it's page 2 in this handout that -- Q. Exhibit I7.1, page 2, the heading of which is: O&M PBR Formula 1987-97. Is that what you're referring to? A. Yes. And, again, what I did is I took the company's data as filed in the prefiled evidence, I accepted the O&M PBR formula as proposed by the company - so it includes this .63 per cent productivity offset - I adjusted for customer growth based on the company's data, it uses the actual Ontario CPI as an inflation index to adjust the cost of service for inflationary tendencies -- pardon me, the O&M expenditures. And what you don't see very clearly in this diagram, but you see much better in the following one which is really just page 3, called Sufficiency/Deficiency 1987-97, and this is the difference now between what would have resulted from applying the formula and the actual O&M budget coming out of the rate case of that year. So this is the difference between the Bauer 1077 dr ex (Warren) hypothetical and the actual, if you wish. And you'll see that in all but one year there are significant differences, and the hypothetical O&M budget derived from the formula is always significantly higher than the budget derived in the rate case. MR. VLAHOS: Sorry, Dr. Bauer, was there a rebasing done on this schedule every three years? THE WITNESS: This one doesn't, this is the next -- the next diagram shows you the rebased one, okay? MR. VLAHOS: Sorry. THE WITNESS: This one doesn't rebase. This would -- we could be in this scenario easily if in three years the company came back to say the best strategy from now on is to repeat the plan as it was. And, as I said before, it has been the preferred negotiating strategy in many jurisdictions where such rebasing took place. So in that case, and given this proviso, you see an accumulated over-earning of $50-million plus. I didn't include the year 1998. If you would include this, you would again move up to a whopping positive deviation between the formula and the actual budget. Now, your argument is, of course, justified and I'm just sort of trying to sort of help us understand the mechanics of these formulas, okay. If now we would assume that we are able to rebase every three years, and if we can create the conditions that allow us to assess the O&M budget meaningfully and, Bauer 1078 dr ex (Warren) you know, if there are there years of no reporting or limited reporting this may not be the case. So this is a very strong -- there are very strong assumptions underlying the second scenario. There are again two diagrams, one called: O&M PBR with Full Rebasing, and then again the second diagram that shows the difference between the O&M formula result and the actuals. And what you see here is that in the 10-year period there would still be accumulated over-earnings of about $7.7-million and you see on page 5 that it varies from year to year and this is largely due to the success, if I should say, of rebasing where deviations that occur after three years are adjusted or controlled for by the rebasing. Again, if we would add the year 1998 you would see a spike upwards, a positive figure in this diagram. What is interesting to see perhaps is that if one would stretch the productivity factor by about .4 percentage points above the current .63, then the accumulated gains and losses would be approximately zero. Now remember, this is based on the incentives of the company working under cost of service regulation, okay; this is not the company working under incentive regulation. And the company has stated repeatedly, and I think we all agree, that the incentives to look for efficiency increases on incentive regulation are higher. So, in that sense, this is really sort of a -- Bauer 1079 dr ex (Warren) for the new world, this is the worst case scenario in a sense. You know, things very likely will be better, the efficiency gains that the company can achieve should likely be higher than this. And now coming back to your initial question, looking at this data, I think that the argument that ratepayers will benefit from the plan as it exists are not very convincing. Actually, it seems that ratepayers and their representatives are not having an opportunity to review data on a more continuous basis and a risk to lose out in the longer run because you may not have the information to rebase at the end of these three-year periods. I should also say though, but perhaps we can come back to that, that there are ways to, in my view, to fix this issue, you know, to come a solution that, indeed, will create benefits to ratepayers and not the uncertainty that we see here. Q. Two final areas, and I'll return to that question in a moment, Dr. Bauer. But just before I do so, there has also been a great deal of discussion in the case about the O&M base and of what is included and what is not included, including a discussion of whether or not there should be a true-up mechanism so called. Do you have any opinion, Dr. Bauer, or any concerns with respect to the O&M base? A. I have very strong concerns, I should say. Bauer 1080 dr ex (Warren) I have emphasized many times before that the success and failure of a PBR plan depends on how well we are able to determine the components that make up the formula, and there are really four key elements here. One is the productivity offset that is used in the formula; the second is the inflation rate that we use to approximate price inflation; the third is the question of adjustment to external factors that we try to capture with this the "Z" factor mechanism; and the third (sic) issue is sometimes overlooked is where do we start from? And one thing is clear, if you start from the wrong base then the entire plan is misled. Now, if you would go back to this initial diagram here, page 1 called Benefits of Indexing, it's very simple to see here if the true initial base were $230-million instead of $240, you make a mistake of $10-million per year. And there's no way that this windfall will ever disappear, you know. So, therefore, considerable attention needs to be dedicated to this starting point. Now, I have also argued in other circumstances and even here that it is justified to use historical data as a starting point, especially if the past practice of rate setting and cost reviews gives us sufficient confidence that the data are meaningful. But the problem that we have here, I think the process in this jurisdiction was meaningful, and I assume sophisticated, but you face great uncertainties as to how the business will look like in the future. Bauer 1081 dr ex (Warren) So while the method of how you determine cost here I think was sophisticated and reliable, the uncertainty is how is the business going to look like. And the data provided by the company are really based on the forward-looking -- they are a forecast, it was probably done some time in 1997, as to how the business would look like. So we have a hypothetical unbundled budget as a basis for a multi-year PBR plan. I think this is really an undesirable scenario. There are several ways how this could be remedied. One would be to say: Well, we accept this hypothetical budget, but we true it up -- the base, we true-up the base once we have the actual data available. This is, in my view, the only way this problem can be avoided. It's difficult to say what the magnitude of this problem is simply because we don't know what the actual budget is. And I think it would really be a very poor regulatory approach to start a PBR plan from a hypothetical budget. Q. Finally, Dr. Bauer, your recommendation as contained in your evidence is that the Board should turn down the targeted PBR proposal in favour of waiting for a more comprehensive one. But I would ask you to assume for the purposes of the following discussion that the Board decides it wants to approve a targeted PBR proposal, what do you believe are the essential changes, if any, which need to be made in the targeted PBR proposal in order to provide, in your Bauer 1082 dr ex (Warren) view, adequate benefits and protection for the ratepayers? A. There are four areas I think that need to be addressed. First of all, in the current plan what happens is essentially we use the status quo and prolong it into the future. So customers -- ratepayers, I should be more precise, ratepayers are, from that perspective, essentially not better off or not worse off than they were in the years underlying this projection. However, we talked about the uncertainties that we face in terms of rebasing, for example, therefore, I tend to believe that in the current environment, or if the plan would go ahead as proposed, the ratepayers are indeed slightly worse off because they, you know, they get the same expectation in terms of what they pay for service but there's more uncertainty as to how well they could defend their interests in future proceedings. This could be remedied, in my view, in a number of ways and this leads me to those four points. First of all, one could envision to include - as indeed many regulatory agencies have done - some type of up-front benefit to ratepayers into the formula. That has been done under several terms, some regulatory commissions have called it "stretch factor", others have called it "consumer dividend". The magnitude of this adjustment which has several rationales - one of them is actually that the Bauer 1083 dr ex (Warren) company in the past worked under a monopoly conditions that, therefore, the company's own record certainly is not a good predictor of the future - it's certainly not of a market standard that one would want to use in such a context. Also that perhaps many of the efficiencies that could be achieved we cannot really pinpoint, certainly not if we use past historical data to forecast productivity gains. The magnitude of this factor has been practically between .5 per cent and 1 per cent in regulatory jurisdictions and this is not just telecommunications, but it's across the board in energy and telecommunications. An alternative to this method that would also assure, rather than just promise future benefits to the ratepayer, would be the establishment of an earnings/sharing mechanism. Now, that has the disadvantage, in my view, that it increases regulatory complexity because, you know, a monitoring process has to be put into place, one has to agree on an earnings/sharing formula and many of these issues do not have straightforward answers. So my concern would be that in this model of an earnings/sharing mechanism the potential economy in terms of regulatory costs would even be worsened. Secondly, in order to provide the information that is necessary to assess the success of the plan, I think it is necessary to establish some reporting Bauer 1084 dr ex (Warren) requirements that allow, at least the Board - not necessarily the public in the forum of a full hearing - but at least the Board to monitor during the plan period what the impacts of the PBR plan are in terms of operating efficiency, in terms of operating costs, on the utility. And quality of service costs and so on would have to be part of such reporting. Thirdly, one of the concerns that I have in the current environment is that in one area it's almost as if the Board -- that the Board is forced to move backwards and that is, the entire area of affiliate transactions. There's really little - at least that's how I see it - that the Board could currently do to avoid, given the new undertakings, that the company engages in affiliate transactions that shift some of these efficiency gains out of the utility to an affiliate. Now, I know that there is a discussion on a code of conduct, and I know that there are concepts in place that one could use to avoid such shifts. I don't see, though, how this will be enforced and monitored on a regular basis, which leads back to the second point, that there needs to be some monitoring in place. Last, but not least, I think we have to think very carefully about the other parts of this plan. For, one, I guess "Z" factors should be defined in a narrow concise way to avoid extended debates. MS. LEA: I'm sorry? THE WITNESS: Debates on, you know, what Bauer 1085 dr ex (Warren) qualifies as "Z" factors, what the impact should be. And I think that the record in other jurisdictions really gives a very clear guideline for how to do this. I mean, it's clear that external tax changes affecting gas distribution companies, regulatory changes affecting specifically gas distribution companies, accounting practices would qualify as "Z" factors. Other things that have been proposed such as litigation costs or stranded costs, I don't see how they would qualify as such adjustments. Likewise, I think the quality of service mechanism, if the plan should be meaningful, should be put as much as possible on auto pilot. The idea of PBR is really to take as many issues as possible out of the daily regulatory discourse, or the annual regulatory discourse. In the case of "Z" factors, you do it by narrowly defining what qualifies, and then only under very unique circumstances could one bring forward another event to the Board. In the case of service quality indicators, the best way to establish such an auto pilot mechanism is to say: Here's my service quality indicator and if you fail to meet it there's a certain consequence. It could be a penalty. If the jurisdiction doesn't allow establishment of such a penalty, it could be folded into the next year's O&M budget by modifying the revenue cap. So there are many ways to treat -- to deal with Bauer 1086 dr ex (Warren) it, but it would be desirable to have an automated process in place, rather than a debate as the issue arises. And, last but not least, I think a good plan, given all the other provisos that I made, would have a plan in place as to how the PBR plan will be reviewed at the end of the plan period, and that would have to have a commitment as to how this rebasing will take place. At least the methods of how this rebasing will happen should be specified. MR. WARREN: Thank you, Dr. Bauer. Those are my questions. THE PRESIDING MEMBER: Which counsel will be questioning Dr. Bauer? MR. THOMPSON: I am a co-sponsor. THE PRESIDING MEMBER: Ms. Lea, Mr. Cass? MR. CASS: That's correct, Madam Chair. THE PRESIDING MEMBER: Ms. De Marco? MS. DeMARCO: I have one quick question. THE PRESIDING MEMBER: Go ahead, Ms. DeMarco. CROSS-EXAMINATION BY MS. DeMARCO: Q. Dr. Bauer, just following up on a question that Mr. Warren asked regarding a comparison between this targeted PBR plan with the existing cost of service regime, specifically looking at whether or not the ratepayers will be better off under this proposal or under existing cost of service, I wonder if you can specifically address the area of service quality and indicate whether the ratepayers would be better off in the area of service Bauer 1087 cr-ex (DeMarco) quality? A. Well, it all depends I would say initially on what service quality target will be specified. If we use the status quo, which I think in certain areas has not been entirely satisfying, then they probably would be worse off. But if, let's assume that it is feasible to specify quality of service in a meaningful way. Then I think what is necessary to assure is that ratepayers are not worse off. Now, that assumes that the past historical level of reliability, quality of service is satisfactory and I cannot speak to that entirely. You all are better experts than I am in that area. In my view, as we envision a more open, more competitive natural gas industry, what will likely happen is that in some areas where there will be competitors who enter the business, we will see a differentiation of service quality. I think that this notion of a predetermined level of service quality that is the same for everybody will gradually vanish in the future because the logical consequence of a more open market environment will be a differentiation of qualities and prices. And what we hopefully will see is that better quality will be available at a premium and perhaps worse quality hopefully at a lower price. What we don't want to see is worse quality at a premium price. Bauer 1088 cr-ex (DeMarco) And I have to say that our examples were largely due to regulatory failure, such observations occurred in the past. Many of the initial PBR plans actually did very poorly in specifying service quality targets and as a result, what we saw is soaring profits and declining service quality. I think that we have all the tools necessary to avoid such a happening. Does that answer your question? MS. DeMARCO: Essentially, yes. Thank you. Those are my questions. THE PRESIDING MEMBER: Thank you, Ms. De Marco. I think we will take a short break before we hear from Ms. Lea. Ten minutes, please. ---Recessed at 3:27 p.m. ---On resuming at 3:46 p.m. THE PRESIDING MEMBER: Ms. Lea. MS. LEA: Thank you, Madam Chair. CROSS-EXAMINATION BY MS. LEA: Q. Dr. Bauer, I would like to attempt to review a few points with you, attempting to organize my thoughts in regards to the four recommendations that you made towards the end of your examination in-chief. I think there were four things you said that need to be fixed if this proposal was to go forward. I wonder if we could start with the first item which was a discussion of up-front benefit to ratepayers. There were a couple of things I wanted to ask you about this. Bauer 1089 cr-ex (Lea) First of all, the company in its evidence has talked about the fact that it will be providing an up-front benefit to ratepayers. And at page 12 of your evidence, at page 12 of your evidence, you indicate that you believe the characterization of what the company is going to do as an up-front benefit to ratepayers to be misleading. This occurs in the last paragraph of page 12. This interpretation is misleading. I wonder if you could just explain to us why that is misleading or what is different about what the company is saying is an up-front benefit and what you're talking about as an up-front benefit? A. Yes. I think what the company refers to when it speaks about an up-front benefit is the fact that ratepayers under this model can - you can correct me if I am wrong - have the certainty that O&M costs will increase at 0.63 per cent lower than the Ontario CPI. Now, if one accepts that certain expectation is a benefit, then one could call this a benefit. In my view, the test as to whether benefit occurs needs to be more stringent and the appropriate test would be, are there compared, you know, to the past record of the company any improvements that consumers will derive? Q. I wonder if you could, when referring to the company, you can use the name Consumers? If you're referring to ratepayers or customers, you might want to use those words. A. Well, given the sort of diversification Bauer 1090 cr-ex (Lea) attempts, I think it becomes increasingly difficult. Ratepayers. I think I follow in Mr. Warren's footsteps and talk about Union -- Q. Yes. He started it. That's true. [Laughter] MR. WARREN: Me? THE WITNESS: Infected with the same virus. MS. LEA: No, no. It's true. You did. MS. LEA: Q. Or you can say small "c" consumers; whatever you like. A. In my view, the company has historically had a certain performance record, and it seems reasonable to assume that this performance record can be extended into the future. It is also the company's own assumption when it uses its own past data to forecast this benchmark; therefore, the benefit would have to be measured as a difference in the world under the PBR plan compared to this longer run trend. And the way the parameters of the PBR formula are currently specified, they do not add an additional benefit to the long-run trend. Indeed, the target is the longer-run trend that you see here, and that's, in my view, a more meaningful way of measuring benefits would be compared to this longer-run trend. So the company promises the longer run trend with certainty. In this very, very limited sense, this certainty can be seen as a slight benefit, but in the more Bauer 1091 cr-ex (Lea) substantive sense of what we can declare a benefit, the plan does not provide such traditional benefits. Ratepayers, if we would for a moment discard all the issues related to uncertainty as to how the rebasing would take place and other things are related to the amount of influence on the process of regulation, ratepayers essentially should be indifferent between cost of service regulation and this plan. And that is, for most regulatory agencies, an insufficient reason to adopt the PBR plan. You can, for example, look into the Boston Gas case again, the Massachusetts Department of Public Utilities has clearly stated if we want to go through the exercise of moving from an established well-known regulatory process to a new one, there need to be positive benefits. Q. Thank you. So as I understand it then, the use of a stretch factor or an earnings sharing mechanism would, in your view, be two alternatives to give the ratepayer some benefit as a result of this plan? A. They would, indeed, provide tangible benefits over and above the longer run trends, and that is in my view the meaningful definition of a benefit. Q. There was one question about the sharing mechanism I didn't understand from your evidence. I wished if we could look at page 4 of your evidence, please, in the last paragraph on that page. Page 4 of your evidence, the last paragraph of that page, and I mean Bauer 1092 cr-ex (Lea) the last partial paragraph. Three lines from the bottom, you have the sentence: 'Key are the specification of a performance target based on data external to the utility and an explicitly or implicitly specified mechanism translating variations from the target inter-rewards or penalties.' I didn't understand what an implicitly specified mechanism would be. I can only think of explicit ones. Can you help? A. Explicit mechanisms would be an earnings sharing process or a process where you calibrate, let's say, next year's productivity offset to the return equity achieved in the previous period. Implicit mechanism I might refer to is, for example, using a productivity measure that is stretched. That is in my view an implicit measure of the amount. Q. Okay. Thank you. That's helpful. Turning to the second of your recommendations which I believe dealt with reporting requirements, early in your examination in-chief, you talked about the need to meaningfully assess true O&M spending or expenses. If this Board wishes to set up a reporting requirement which will allow it to meaningfully assess true O&M expenses in order to, first, effectively rebase and, second, make sure that the company is not over-earning during the time period, can you tell me what Bauer 1093 cr-ex (Lea) information specifically would be needed by the Board to make that assessment? For example, would we need the level of detail of O&M spending by line item? What kind of thing would we need? A. I think two aspects need to be weighed here. One is certainly the question of administrative simplicity of the PBR plan which would speak for more aggregated types of reporting. The other aspect, though, is the question of what approach would facilitate the learning process that is necessary to learn to use PBR techniques, and I think it's interesting to see what other regulatory agencies again have done. Those who try to come up with meaningful PBR plans have in general established relatively detailed reporting requirements to allow an annual evaluation of the process. In other words, in order to create the information necessary to understand what the implications of PBR are, they have required relatively detailed reporting. For example, the British Columbia Utilities Commission is one of those. I think the FCC was mentioned earlier today. The case of the U.K. is interesting and illustrating in the sense that initially the regulators in the U.K. thought they could bypass detailed reporting requirements, and as they saw in the first years of the Bauer 1094 cr-ex (Lea) PBR plans how undesired results happen, such as lower service quality/skyrocketing profits, they gradually reintroduced reporting requirements to the point essentially where they demand financial information, quantitative information that is compatible I should say to the information submitted under rate of return or cost of service regulation. My view is that in the initial phase of the PBR experiment -- let me use the term "experiment" here because I think all the stakeholders involved need to learn how to live in this new world in several ways because, on one hand, we need to learn how to give up micro-management of the process. This is, in my view, incompatible with PBR. On the other hand, one has to understand the different implications for risk for costs and benefits of a PBR plan. And in this initial phase, I think more detailed information reporting seems justified. In my view, it will be appropriate to essentially maintain the existing information filing requirements that this, you know, on a line-by-line basis. But I wanted to highlight this straight off between administrative simplicity on the one hand and the meaningful learning experience, how to effectively use PBR on the other hand. Q. And is an annual filing of this level of detail something that is generally used? A. It is not used throughout, and honestly I Bauer 1095 cr-ex (Lea) need to say that there are regulatory agencies who have essentially abandoned information filing requirement, too. So we see very extreme cases where regulators have decided, 'Let's just abandon all information filing requirement.' That is mostly the case, though, in telecommunications where there's a very reasonable expectation that after a short transition period of three, four, perhaps five years, this market segment will be competitive and therefore regulation will be not necessary, at all. So we see a spectrum of approaches. In my view, the commissions who have carefully thought through how they want to introduce PBR, these commissions have tended to impose stringent information reporting requirements. Q. Thank you. I would like to turn for a moment to your discussion of "Z" factors and service quality indicators, and there is one thing that I need to clarify which may seem like an extremely small and fussy point, but I would just like to understand it clearly. At the top of page 15 of your evidence, in the first partial paragraph that occurs, in the third line of that paragraph, third line from the top of the page, you begin to quote the company I believe from an interrogatory. That set of quotation marks continues to the end of the paragraph. I don't -- do I understand correctly that the Bauer 1096 cr-ex (Lea) last sentence of that paragraph is your opinion or the company's opinion? Should those quotes be moved back? A. It may be as the result of a fuzzy mind too, so I have to cross-check again. Q. I'm sorry to put you to this trouble. I just wanted to understand whether the last sentence was your opinion or not. A. I think this is, as it is taken, is a quotation from the company. So your question is: Would I agree with the statement? Is that what it is, or it's more technical in nature -- ---Off the record discussion MS. LEA: So the quotation marks are in the proper place. Thank you very much, Dr. Bauer. THE WITNESS: Yes, this is really... MS. LEA: Mr. Warren has answered my question and so have you. Thank you. Q. So on page 16 of your evidence then, in the last paragraph dealing with service quality indicators, you talk about several individual components being integrated into an overall index and a weighting scheme being applied. Do you see that such a scheme or an index could be derived from the SQIs that we have in this application or do you think different SQIs need to be used? A. The five that are proposed should certainly be part of this. In my view there are two aspects missing Bauer 1097 cr-ex (Lea) and I'll point these out. One would be an indicator of employee safety. Q. Employee safety? A. Employee safety. Q. Yes. A. And the other item that is frequently included and, in my view, would be very valuable and really rounds the picture of service quality would be independent customer/ratepayer surveys. So, in other words, an independent assessment of whether the expectations of ratepayers have been met in the new environment. Q. Thank you. The fourth point that you made towards the end of your evidence was a commitment to the method of rebasing or commitment to rebasing. Can you be more specific and suggest what the Board should order if it were concerned about ensuring that rebasing would be done and that it would be done right? A. In my view there are two generic ways of doing it. One would be to determine the method of rebasing. That would mean that when the plan is implemented and the order is issued, there is a clear description of the process that will be used. It could simply say: We used established rate case methods and procedures in rebasing the O&M margin. That avoids one of the two difficulties that I Bauer 1098 cr-ex (Lea) mentioned; that is, towards the end of the three-year period the company would argue that in the coming years the situation will be different and, therefore, the basic cost data of the plan do not have to be reviewed. What needs to be reviewed is only whether inflation rate was appropriate and the productivity factor was correctly specified. As I said before, I'm not making this up as a hypothetical scenario. This is indeed was the negotiating strategy that many companies had in those cases where we see a renegotiation of a PBR plan. That is something that anybody who is working under the framework of a manager I should say, to their defence, would do. It's not evil. It's the incentive system of a capitalist enterprise system. The second issue now is not really covered if we agree on the method, and that's the issue of how would the Board be able to detect, let's say, shifting of costs into the utility and -- or how would the Board be able to detect the substitution problem that was discussed earlier between capital expenditures and O&M expenditures? And agreeing on the method alone may not be sufficient to actually ensure that there will be ratepayer benefits at the end of the three years. So another way to handle this problem would be to say: Well, let's take the accumulated benefits to the company and a portion of these accumulated benefits then would be translated into reduced O&M budget. Then you Bauer 1099 cr-ex (Lea) don't just agree on a method, but really on the substance. In this case, you would essentially determine today, what the Board would determine today, how the overall efficiency gains that can be achieved under the PBR plan would be split between ratepayers and shareholders. That, in essence, then, is really a substitute for an earnings-sharing process or a consumer dividend method. Q. Thank you. My last question, Dr. Bauer, is if all of the things that you have mentioned were done, if this proposal that's before the Board were fixed in the fashion that you've described, to the extent that you have described, do you believe that there would be benefits to the ratepayers, would there still be benefits to Consumers, that is the company's shareholders, and would there be some benefit to this Board in terms of reduced regulatory burden or not? So would those three groups or individual -- agencies, whatever, see benefits or not under this proposal fixed in the manner that you have described? A. If these modifications would be executed, it is my view that all stakeholders would be better off in the modified plan, if it is as -- Q. As compared to? A. As compared to traditional cost of service regulation. Bauer 1100 cr-ex (Lea) If it is intended to be a transitory plan to a comprehensive plan, okay. If this is a model that is envisioned as a prolonged model over an extended period of time, my judgment would be that the gains are marginal. So as a transition plan, I think it would be a pragmatic platform compromise to say: Here is how we bridge the gap from where we are and where we want to be in terms of a comprehensive plan. We can do it in a way that will save us time administering the details of the plan if we can agree on an appropriate specification of such "Z" factors, if we can agree on service quality indicators that are essentially on auto pilot, and if we can build into the formula upfront some tangible savings to ratepayers. Now, given, you know, my own sort of preliminary analysis of the data and the company's own testimony, it is very likely that the company also will benefit from such an approach because there seemed to be productivity research that goes beyond the .63 plus a stretch factor. And the experience in other jurisdictions also shows that to a certain degree the amount of efficiency research companies can materialize is indigenous to what "X" factor is set. So, in other words, if you say 1.5 per cent, define 1.5 per cent; if you say 2.5 per cent, define 2.5 per cent. To a certain degree I think, you know, we have to acknowledge that the effort of management will depend on the productivity also that is set and there is some Bauer 1101 cr-ex (Lea) flexibility in increasing it. MS. LEA: Thank you very much, Dr. Bauer. Thank you. Those are my questions. THE PRESIDING MEMBER: Thank you, Ms. Lea. Mr. Cass? MR. CASS: Madam Chair, thank you. Given the time of day, I think the Board and all participants will be happy to hear that in light of Dr. Bauer's examination-in-chief I have very, very few questions left for my cross-examination. MR. WARREN: You don't need to ask any. (Laughter) MR. CASS: Very few. CROSS-EXAMINATION BY MR. CASS: Q. I just have to ask you, Dr. Bauer, in light of what you said in examination-in-chief, have you been able to follow the record of this hearing in terms of the commitments that the company has been making insofar as its proposal is concerned? A. As far as transcripts were available to me, and I have not seen the most recent ones. Q. Okay. The record will speak for itself. Thank you. Just a point of clarification on the Boston Gas decision which you've referred to a couple of times. I don't pretend to be an expert in this, and correct me if I am wrong, but I understand that that decision was appealed and aspects of it were overturned on appeal including the Bauer 1102 cr-ex (Cass) productivity factor being reduced. Do you know whether that's correct? A. It's correct that the decision was appealed and there were slight modifications made. Q. I think there may even be a further appeal from it? Ongoing? A. There is -- I think there is parts -- most of the case is settled. There are -- some of these issues are contested. Q. Thank you. And then just -- A. Also, let me just highlight that this is a very different model because they were talking about the price cap plan and many issues arise in the context of a price cap plan that we didn't even touch here. Q. Thank you, Dr. Bauer. Then just a couple of quick questions about your Exhibit I7.1 which you introduced today. I just want to be totally sure, I believe you've indicated this more than once, but on your presentation, in particular pages 4 and 5, when you refer to full rebasing, that's rebasing after three years? A. It's after three years. I could probably deliver the table that underlies these graph, if that's desired. Q. That would be very kind of you, Dr. Bauer. A. The rebasing in this case takes place -- we are starting in 1987, the rebasing takes place in 1990. So in 1990 -- the former like O&M values rebased Bauer 1103 cr-ex (Cass) to the actual of 1990 that was historically shown in the rate case, it takes place in 1993 and then in '96. I would be pleased to do that. I can deliver that in the near future. Q. If you could do that I would appreciate it. MS. LEA: Thank you. Undertaking 7.1. ---UNDERTAKING H7.1: Dr. Bauer undertake to provide the tables underlying Exhibit No. I7.1. MR. CASS: Q. Just one final question, Dr. Bauer, very straightforward I think. On page 5, I think it's probably obvious, but with the three-year rebasing and -- sorry, page 5 of Exhibit I7.1. With the three-year rebasing and the work that you have done in this exhibit, we can see that had a different starting point been chosen, for example, if the analysis had started in the early 1990s, in fact, the shareholder would have suffered under the PBR proposal, correct? A. It is not correct insofar as 1998 is missing and, as I indicated, '98 you would have a big spike the other way. So what you see is -- there is clearly a difference between the first half and the second half. Although, if we would have more recent data our judgment would be more balanced again. Q. All right. But you haven't -- A. No, but what it indicates though is the fact that, indeed, if your productivity development is not Bauer 1104 cr-ex (Cass) smooth from year to year but there are variations, there can be years when it goes to the advantage of the consumer -- to the ratepayer or to the advantage of the shareholder. Q. All right. So, of course, your choice of the starting point and the time period would be important in looking at the results? A. That goes without saying, although in the current model that you propose the key -- parts of the formula are really longer on trends. So this is essentially almost like the stock market. In the longer run the deviations one way or the other will even each other out. And what is important is really sort of the accumulated -- the net effect of the full plan and it's -- you know, just for the sake of the argument, it's very strongly biased towards the shareholder, would be just -- you know, except no rebasing scenario. It is 7.7-million in favour of the entire period for the shareholder if we accept the full rebasing scenario. And, as I indicated, if we would add a stretch factor, still accepting the notion that we are working under a cost of service incentive structure, by a stretch factor of about .4 per cent would make the plan neutral in the sense that neither, over the period that I looked at here, neither the shareholder nor the ratepayer would lose or win. This is predominantly related to sort of Bauer 1105 cr-ex (Cass) illustrate the mechanics of the plans so that we don't just talk in hypotheses, but we see what the implications are. MR. CASS: Thank you, Madam Chair. That's my cross-examination. THE PRESIDING MEMBER: Thank you, Mr. Cass. Very brief and to the point. The Board has some questions. Mr. Vlahos? MR. VLAHOS: Thank you, Madam Chair. EXAMINATION BY MR. VLAHOS: Q. Dr. Bauer, the last undertaking you have - you have what? - you have committed to provide the back-up data to page 4 of I7.1? A. I think I have committed to provide the numbers that underlie pages 2, 3, 4 and 5. Q. Oh, all of them? A. All of them, yes. Q. Okay, thank you. A. But you'll see this does -- Q. I was going to ask you for all of them. A. Okay. Q. Okay. And before I leave the schedules, if you turn to page 1, let me understand this. The amount of $19.80-million, that is the benefits to the shareholder, you say, for the first term of the PBR, the three-year term? A. Accepting the underlying assumptions such as Bauer 1106 ex (Vlahos) 2 per cent inflation rate, the ability of the company to increase productivity, indeed, at a rate of 2 per cent a year, and accepting the .63 per cent inflation offset factor in the formula. Also, for the sake of simplicity, I have assumed that the customer bases will not change in the scenario. Q. That's fine. A. So you can plug in all these variables and then go through different scenarios. Q. I'm not looking at -- A. But the specific number is really dependent on the specific -- Q. No, I understand that, Doctor. I'm not looking for what makes up the 19.8 but, rather, it is a positive number and I guess my question is: If you can just visualize with me a PBR where it has a life of three years, okay, or you can think of tranches of three years, and there is a net benefit to the shareholder for the first three years of this plan, and assume for a minute that nothing happens after that, we go back to cost of service, if you like, some other world, okay? Now, to the extent that the O&M base would be lowered somehow, whether it's through monitoring, through a cost of service detailed review, okay - and I won't ask you questions on that - but if you accept with me that the O&M will be lower because of those benefits that have been accrued to the shareholder, then the ratepayer would have some benefit from the third year forth; right? Bauer 1107 ex (Vlahos) A. If I accept this, and I indicated what the concerns are -- Q. No, I understand that and I'll come back to that. A. I accept this. Q. But if I want to calculate this and compare the benefits to the shareholder versus the ratepayer, then I would have to look -- the benefits that would accrue year after year after year, starting year four, and to do a proper comparison I have to do some kind of discounting to year -- I guess to year zero, compare the 19.8 - and the 19.8 is not discounted it's just the accumulation of three years, but you can appreciate where I'm going. But if you want to do a comparison of the benefits of a PBR mechanism, you'll have to assume that -- you'll have to incorporate the savings that may accrue to the ratepayer for the remaining life of the ratepayer paying, not being a the customer of the utility. Are you with me on that? A. Yes. Yes, I agree. I agree. Q. Okay. So what the question then is: What method would be applied in year three to rebase the O&M for year four onward, and I believe that is where you have some concerns? A. Right. The benefits that you mention in the forward-looking basis will only accrue if, indeed, you can rebase to that extent. So, therefore, we need, (a) the information that Bauer 1108 ex (Vlahos) allows us to meaningfully rebase; and, (b) we need to agree on a method to rebase. And if we leave this to a future negotiation, my concern would be that we'll change the method and the base as we go. Q. I understand that. A. And the other thing that one has to keep in mind is, on the forward-looking basis, perhaps what also would happen, accepting your scenario for the time being, if the development is three years, indeed, there's a better picture of how efficient the company can be. In other words, it reveals efficiencies that it would not reveal under cost of service regulation. There is really no reason to assume that these efficiencies could not be maintained into the future as well. So if we start to discount these future benefit streams, we also will have to look at the benefits to the -- the continued benefits to the shareholders. We would go through the same cycle again, there may be -- you understand what I'm saying? Q. That's what I said. Think of it in tranches; the first three years and then, you know, the second three-year term. If it's true for the first term of three years, then it would be true also for the second three-year term, and so and so on. So if you're with me for a minute then. If you assume that this panel is very -- it's the best you can Bauer 1109 ex (Vlahos) have and we can identify every inefficiency in year four, then those benefits to the ratepayers would be accruing to the ratepayers? A. Right, I agree. If we could do that, that's correct. Q. Okay. So now the question is how we get -- what method do we use to rebase for year four? And that's where I would like to ask your understanding, and I must say that I'm not clear yet - and I'm saying that for the benefit of the company as well - as to what is the proposal, absent a comprehensive PBR, to replace this three-year limited PBR; what is the recommendation for rebasing, and is it going to be whatever the actuals will be for year three, or is it going to be a result of detailed review "a la" cost of service, okay? And that's where I'm not clear and I don't know if you have any impression from the evidence. If you do, help me. A. I think I can best help you out by sharing with you what I think the appropriate method would be. I mean, in my view -- and, as I said before, I haven't had a chance simply because I didn't receive some of the transcripts and maybe in the last few ones there may be some indication of how this process, in the view of the company, should happen, but I didn't see any of that in any of the prefiled evidence or the interrogatory answers or the transcripts that I saw. Bauer 1110 ex (Vlahos) One has to keep in mind when going through this rebasing process that the company, Consumers, still operates here in a monopoly environment. Now, it would be justified to use actuals, in my view, if we could be sufficiently confident that these actuals reflect a market standard, but we can't because the company continues to operate in a monopoly environment. So I don't think there is any way out of scrutinizing or auditing these actuals to make sure that they are not hiding inefficiencies. And I think there are certain parts of the process where one probably could economize, compared to a full-fledged detailed rate case. And it may not be necessary to go through the full details, let's say, of a Phase 2 in terms of detailed rates design. I'm somewhat tentative on it, I'd have to probably think more carefully about this. But here's where I'm coming from. What we want to assure in the review are two things. First of all, we need to assure that the level of O&M expenditures in our narrow scenario here, indeed, reflects prudent operations, that refers to the level of expenditures. Secondly, I think it is important to assure that in the process, in the three years we haven't incurred any changes in the cost structure of the company that are not reflected in the rates. Now, I think that we can simplify, as we probably don't have to look at every single rate component, rate Bauer 1111 ex (Vlahos) element in detail. What would suffice is what is sometimes called, has been called category cost reviews. So that you would pool similar customers in one category and you assure that as a group they cover their costs, and this would help us understand or eliminate the risk of cross-subsidization between those groups. And these things -- as costs shift, as demand patterns shift, these things almost certainly will happen. But it is possible, in my view, to envision a somewhat simplified type of cost review in this rebasing process, based on some of the tools that we know from a rate case. Q. Okay. You mentioned one of your concerns is the affiliate transactions, to what extent some of the efficiencies that can be had in the utility will be exported, if you like, to the affiliated companies. Now, it's my impression, Doctor, that this problem is, this issue is problematic based on your testimony, if you have a situation where rebasing is done on the basis of the 30 years actuals, and perhaps less so if it's going to be a more, a fuller detailed review. Did I read that correctly? A. That is correct. If the information is available and we, indeed, conduct a review of the prudency of expenditures, then the risk is mitigated. I think that the code of conduct will play an important role. I don't see at this point how it will be monitored. The potential risk could be the following, that a Bauer 1112 ex (Vlahos) stream of, let's say, outsourcing starts in year one of the plan, and then in year three at the time of rebasing you see costs per unit of service delivered but you cannot really go back three years and compare it to the market standard, so I think unless there is sort of an ongoing reporting process and monitoring process in place, you may run into situations where it's impossible to assess whether an expenditure is prudent or not. And now this faces the difficulty anyway to find for many expenditures market standards that even in an ongoing monitoring process may be difficult to do. And the answer, unfortunately, I think to this problem is a continued reporting process, a continued monitoring process which reduces some of the administrative benefits that would otherwise be associated with the rate case. Now, let me also mention one thing, that the concern about such affiliate transactions is mitigated in a price cap scenario; that it is particularly high in a revenue cap environment. The price cap model has built-in safeguards that mitigate, if not eliminate the risk. I mean, one has to in detail look at how well the price cap is designed and this all is depending on a PBR design that is efficient and not biased. But assuming they're able to do this, structurally the price cap model does not raise this issue as much as a revenue cap model. Bauer 1113 ex (Vlahos) Q. Thank you, Doctor. Just one last question in this same vein. You gave us a number -- you actually numbered them, the things that we must fix if we are going to go ahead with a limited PBR plan. And although you spoke of earlier at the beginning of your testimony about the gold plating issue, you didn't have any advice as to how we can fix that if it needs fixing for the duration of the plan. Do you have anything to offer on that or does it fall in the same category as the affiliate transactions that we spoke of? A. In my view, what can be done in the framework of this limited plan is to use the tools that are available from traditional rate cases in monitoring the capital expenses. And that may be imperfect, but I think it's probably the best one can do at this point. Now, of course, going to a comprehensive plan would be a more desirable way to go and that may be the longer term goal. Close, close scrutiny, I think, is the best answer in this case, but it's imperfect. To a certain degree, I think Dr. Norsworthy's proposal to actually try to incorporate some of the capital productivities into the "X" factor, you know, would take care of some of this issue because it would reduce this incentive to shift towards the capital side. But if we accept that this is a transitory plan, if some of the other modifications that I have recommended Bauer 1114 ex (Vlahos) will be put into place, I think for the transition period, one can live with a model where the capital side is monitored in the traditional way. Q. Thank you, Doctor. I did ask -- one of my questions earlier was what was your impression of the company's proposal, whether rebasing means taking the 30-year actuals and moving them forward or was it the subject of a detailed review. And did you have an impression as to which way the company was proposing or not? A. No, not that I'm aware of. MR. VLAHOS: All right. Thank you, Doctor. Thank you, Madam Chair. THE PRESIDING MEMBER: Thank you. Dr. Higgin? DR. HIGGIN: Thank you. EXAMINATION BY DR. HIGGIN: Q. Dr. Bauer, I wonder if we could sort of follow on with this question of rebasing sort of combined a bit with plan review at the end of Year 3. And perhaps to help with some questions, if somebody could provide you with a copy of the Board's decision in EBRO 497 and specifically ask you to look at the structure of the O&M budget in that decision. That appears at page 35 of the document. Do you see the table there? A. Yes. Q. Now, just, we're not really worried about the Bauer 1115 ex (Higgin) numbers. The one on the right is the estimate for the bridge year and this was the company's proposed budget for -- and so the actual Board-approved would be different from this. But I'm really looking at the structure issue now. What do you do in terms of both either in-plan or at the end of the plan - let's start with the end of the plan - when there are structural changes in the basket of services that the company is providing that may not have been dealt with as a "Z" factor. For example, they exit the merchant function; for example, they exit NGV. You know, there are structural changes that are going to occur that maybe should be off-ramps, that maybe shouldn't be, or they maybe should be dealt with in some way. But leaving aside now the idea of off-ramps, what do you do when you review the plan and there's some fundamental structural changes in the basket of services that is being provided that have occurred between Year 1 and Year 3? A. Right. Some of the examples you mentioned I think deserve being treated as an external adjustment factor. For example, if the company were to decide to quit to outsource its call centre and perhaps marketers then bill customers directly for their related services; in other words, there would be, indeed, a savings in the O&M, then this would have to be corrected for. Bauer 1116 ex (Higgin) Generally, any type of elimination of a service that then would be billed directly to customers, ratepayers, I should say to be more specific again, would have to be reflected in such a satisfactory adjustment. And, you know, that would certainly qualify as such an external change. And it may be worthwhile looking into, for example, BC Gas's PBR plan where provisions for such scenarios are very clearly spelled out and, for example, you know a company eliminates certain activities from its own operations. In the case of BC Gas, it is allowed to account for efficiency gains to meet the productivity target, but it is forced to eliminate the O&M component corresponding to an outsourced service for the level of the O&M revenue cap, and I think rules for this can be found. The other question is - and this was why I insisted that at the point of rebasing, one has to at least make sure that customer groups, customer categories do not cross-subsidize each other because such shifts can really happen because of changes in the market demand, changes in the way we bill for delivery services that cannot be anticipated at the point in time when the plan is put into place. There are really sort of two types of structural shifts. One is that the company as an organization charges dramatically. This would have to be adjusted through a "Z" factor, although in my view "Z" factor Bauer 1117 ex (Higgin) adjustments have limits. I think they're more appropriate for more marginal changes than for fundamental changes. And perhaps if such fundamental changes are anticipated at this point, the prudent decision may be to wait until these fundamental changes have occurred and start the PBR plan then, you know, when there's less uncertainty affecting the level of the revenue cap. Q. Okay. Just slightly following up on that, another effect would be that the mix of inputs might change due to these structural changes. And so that by the end of three years, the share of labour and materials, for example, in the company's limited formula may very well have changed due to these structural changes. And should you then try to look at that again from fundamentals or should you just be doing a historic review of the productivity factor and giving it your best shot for the forward period when you are looking at reviewing the plan? A. Well, the changes in the input structure certainly will be reflected in the productivity data of the company. Now, this again-- Q. There's a lag usually. A. --shows the limit of using just a partial productivity measure and, again, highlights the advantage of moving to a broader based PBR plan because, on the broader based plan, what you could do, what one would have to do is essentially move, go to a moving average Bauer 1118 ex (Higgin) calculation of productivity changes. So you will take account of the most recent year's, the most recent experience and let go of the oldest three or five years, whatever. In the current proposal, if it is then translated into a comprehensive plan, perhaps the issue can be ignored based on the assumption that the movement is towards a comprehensive plan anyway. If, however, during these three years there are significant shifts to increase, to inflate, modify the capital bases relative to other inputs, I think then one would have to review it to redesign the basic parameters for the new comprehensive plan. Q. So just coming then to the rebasing issue, a simplistic approach would be to say whatever the actual O&M was in Year 3 of the plan, that would form the base for Year 4, if the plan was to continue. What I'm trying to explore with you, two things; whether there are adjustments that maybe should be made at that point and, if so, how do you get at those, or perhaps another related area is what other aspects of the fundamentals of the plan of the formula, assuming you were continuing with a targeted, would you need to review and reset? And that's what I'm really trying to get. A. Mm-hmm, yes. Q. What would be the scope of that end of Year 3 review? A. Right. In my view, let's assume the same Bauer 1119 ex (Higgin) basic plan structure would remain in place. Then the O&M base rate would have to be rebuilt in my view more or less from scratch or perhaps starting from the actuals with modifications to reflect perhaps structural shifts that you referred to, and I think that the established processes of information reporting and auditing would probably suffice to do that. The big issue, I think, that will need to be addressed is the question, how -- what weight should the actual return equity the company was able to achieve during the planning period, how strong should that be factored into the review? And there are really two different schools. One is to say it should have no impact, at all. We just leave it out because part of the incentive plan is, you know, to allow companies to be more efficient in its reward. It can earn higher profits. The other school, which I think I can identify with more strongly, at least in an environment of continued monopoly. If you're in a transition to competition, I think I belong into the first camp. I think we can leave profits out of the picture. But if we are in a continued monopoly situation, we have to look at the profitability of the company and perhaps make an adjustment based on the understanding that increased profits during your planning periods are an aggregate indication of the ex-inefficiencies that we cannot detect otherwise. Bauer 1120 ex (Higgin) So in my view, there should be some adjustment to bring the cost structure of the company back to a normal return on equity at the point of review. DR. HIGGIN: Okay. Thank you. Thank you for your answers. Thank you. THE PRESIDING MEMBER: I think Mr. Vlahos has one more follow-up question, please. FURTHER EXAMINATION BY MR. VLAHOS: Q. Dr. Bauer, I apologize. There was one question that I had noted and I neglected to ask you. When you talked about what you call the up-front benefit and then I put in brackets Consumers' dividend and a stretch factor and, or a sharing formula. Now, those are, I guess, mutually exclusive, are they? A. Not necessarily. Q. Not necessarily. A. And they were actually used in combination in the telecommunications case. I personally think that one should take the PBR idea serious in the sense that we should try to keep plans as simple as possible, and to combine an up-front dividend and an earnings sharing mechanism may be an unnecessary complication. It is feasible. I mean, the interesting thing about PBR is that there are many different choices that regulatory commissions and utilities have made. And I Bauer 1121 ex (Vlahos) think we are, in terms of PBR, we're at a point where perhaps rate of return regulation was in the 1920s or 1930s when many experiments took place, but there was essentially very little agreement on really what assumptions we should take. You can look at rate of return regulation as a method that grew gradually over time and where one of the most important things that happened in the early decades was that we agreed, that the stakeholders agreed, at least in most areas on how to approach issues, how to determine, you know, the cost of capital for example. And in my view, the discussion on productivity measurement really illustrates this very clearly because what we have on the record, in my view, are two approaches; one taking the very generous role for the company and the other one taking very stringent assumptions that perhaps are more in the interest of ratepayers. The truth very likely is somewhere in between, but it will take a while. Both sort of set of data really can be justified in their own arguments, and what is missing at this point is an agreement on how exactly we're going to do it. And likewise, when we are talking about earnings sharing, we haven't really found the convention that everybody applies. There is a general tendency, I should say, to not use earnings sharing mechanisms in the long run. They can Bauer 1122 ex (Vlahos) be an appropriate risk insurance, if you wish, in transition periods because that's exactly what an earnings sharing plan does. It mitigates the risk for the company on the one hand; it mitigates the risk for the ratepayer on the other hand. And whenever unexpected things happen, there are contingency plans in place, you know, how the costs and benefits will be split. That may be the risk insurance that is desirable in a transition period. Q. Yes. I guess one of the criteria that was adopted by the company in my questions to Mr. Grant some days ago when I asked him from a regulatory efficiency point of view which one would be easier; a stretch factor or a sharing arrangement, he did agree that it was a stretch factor, and for obvious reasons because sharing formula, you have to first determine what's the share. A. There are many other issues here, absolutely correct. Q. But help me with this so-called consumer dividend versus a stretch factor. I think in consumers dividend you spoke of that at the end of the three year-term you determine how much the O&M has declined from the original base. I'm sorry, I didn't follow that part. A. The idea of a consumer dividend or a stretch factor essentially refers to the same concept, to take the productivity factor as determined based on the model, and I have deposited my reservations to using a historical based productivity factor in my evidence, and then Bauer 1123 ex (Vlahos) acknowledge that because the company was able to operate on a monopoly environment that there are many hidden inefficiencies. The economic literature has created the term ex-inefficiency because we don't know what they are, but we know they exist. Acknowledging the fact that there was a monopoly environment and such ex-inefficiencies, then the logical conclusion is let's stretch the productivity factor by .5 percentage points, 1 per cent. That needs to be weighed in the light of the specific circumstances of a case. One option that I mentioned that one could at the end of the three-year period say: What is the accumulated value of the benefit to the shareholders, and then agree on some kind of link between these accumulated benefits and to rebasing. That is not commonly discussed under the term stretch factor or consumer dividend. Q. It's not? A. No, it's not. Q. What's the term you would -- A. It's my laying out to you the options that exist. Q. You have to give it a name. Okay. Thank you. A. Dr. Bauer's proposal, if you wish. MR. VLAHOS: Thank you, Doctor. THE WITNESS: Thank you. Bauer 1124 ex (Presiding Member) EXAMINATION BY THE PRESIDING MEMBER: Q. Can I follow up just a little on that. Your number .5 to 1 per cent, that's a kind of hand waving number or where does that come from? Could it as easily have been 1 to 2 per cent, et cetera? A. This comes really from examples in other jurisdictions. And I think I highlighted in my own paper that our research insights as to the appropriate specific value for stretch factors is, unfortunately, less developed than our understanding of how we measure productivity in other things. So there is some discretion involved. I will be very open about this. But the values that other commissions have chosen tend to be in this range, between .5 and 1 per cent, and they are based on probably a -- I would say an educated guess as to how high this productivity gains could be. I've tried to give you a little bit more solid information by trying to point out what stretch factor would be required to look at this ten-year period and say at the end of the period we are in a worse situation; neither the shareholders have benefited, nor have the ratepayers suffered or benefited. This is a good measure because what I really assumed in this scenario is that we were as well off in the PRB scenario as we were on the cost of service regulation. I also said that very likely, given the different incentive structure on the PRB, that's a very pessimistic Bauer 1125 ex (Presiding Member) scenario, but let's for the sake of argument just assume this, then given the data is available and not taking into account the year 1998, the stretch factor would take us to a level where neither the shareholders benefit nor ratepayers suffer is probably if the range of .4 per cent. So that's based on the historical record that exists. So probably productivity offset factor at around 1 per cent over the 10-year period would lead me to believe that the accumulated net benefits are zero. So neither shareholders -- that's just the one steppingstone in thinking about such a stretch factor because what it does here is provide some empirical evidence to how high it would have to be if we applied the same formula to the past record so that the company wouldn't really benefit from it. If one subscribes to the argument that under PRB the company can do better than in the past, then one would say: Well, now we have to add a few more tenths of a percentage point likely. The second part is probably more difficult to conduct based on empirical and research traditions. It's more based on weighing the pros and cons of different features of a case. Q. And, of course, as has been pointed out, your 10-year calculation would be very different if you had started it in a different year. So that is some empirical evidence as to the size of it, but it's not very much more persuasive than my Bauer 1126 ex (Presiding Member) losses on the stock market, for example, are persuasive. The stock market never goes up. A. It isn't in -- what we have to agree upon is what is a meaningful time horizon to make such assessments. I think it is not justified to say that two years or three years is such a meaningful time horizon because we'd like to mimic are longer-run trends. So I think the mistake by just picking, let's say, the last three years could be very significant because we don't know how the next few years would look like. Even in my own calculation here, the year 1998 isn't in here. I will put it in my table, but actually the positive deviation in the year 1998 is a whopping $22-million. That would reverse the picture that you seem to see in the last few years on page 5 which reversed it significantly. That just shows, you know, how easily we make a mistake by just looking at very short time periods. Q. Can I just ask one more question related to that. When you say .5 to 1 per cent and we're talking about a proposed figure of .63, if the company in other examples that you know of had been proposing a different number, how dependent is that .5 to 1 on sort of bracketing what the company has proposed? Would it be much changed by a situation in which the company were proposing that they could produce much Bauer 1127 ex (Presiding Member) better productivity gains? Or less? Would it be inversely proportioned? A. The big question is here is what is our reference point and the conceptual weaknesses in using the company's own historical record abound. I mean, I have been very outspoken about this. So I would say the following: If compared to a benchmark, which likely would have to be an industry benchmark, we can say what the company promises to do is above the industry benchmark, then the need for such a stretch factor or consumer dividend may not be -- may not exist. If, on the other hand, the company's proposal is significantly below that, then it may exist. But that's just one factor. The other fact -- so the relationship that you point out, that if the promised productivity offset factor in relationship to general practices is high, I think then the need for a consumer dividend decrease and vice versa. Q. Okay. One question about the "Z" factors. I think you were saying it's important to have a narrow definition of the "Z" factors so there isn't ongoing debate as to what constitutes a "Z" factor. But that wouldn't obviate the possibility that you would define certain things as "Z" factors and still leave it open for the company to come up with some unusual thing-- A. No, absolutely. Bauer 1128 ex (Presiding Member) Q. --as a "Z" factor? A. That would be, in my view, the ideal solution to say: Well, here are certain predefined ones. We don't have to go lengthy debates to agree that these qualify and tax changes, accounting changes, regulatory changes are clear candidates for such a treatment and then there could be a general category where you say: On the specific unexpected circumstances there can be an appeals process to the Board. Now, what worries me somewhat is that it seems that this appeals process, given your rules here in Ontario, doesn't seem to be symmetrical because there could be situations, for example, where downward "Z" factor adjustments were necessary but the company obviously would not come forward to the Board and claim it. In other jurisdiction where I have worked, consumers, ratepayers, have the opportunity to petition the regulatory agency to trigger such adjustments. I'm not sure whether this opportunity exists it. I think it would be necessary that there is some symmetry to avoid a situation like this. You know, you can look at my historical data here in that light because the only way to get from the formulaic value of the operations and maintenance to the actual ones would be to have continuous downward "Z" factor adjustments. At least in a majority of the years. THE PRESIDING MEMBER: Thank you very much. Bauer 1129 re-ex (Warren) Mr. Warren? MR. WARREN: Yes, thanks, Madam Chair. RE-EXAMINATION BY MR. WARREN: Q. At the risk of -- in an attempt to ensure that the record is clear or at least in my head is clear, and with the danger that I may muddy the waters, I want to go back, if I can, to the exchanges that you had with Ms. Lea and various members of the Board on the stretch factor and where we begin. Let me ask you if I've understood it correctly, Dr. Bauer. The stretch factor -- is it the case that the stretch factor is added to whatever the Board determines is the appropriate productivity number? A. That is correct. As an acknowledgment for these ex-inefficiencies that we cannot pinpoint and not quantify but we know that they exist. Q. Now, one of the issues that came up in the discussion with Ms. Lea of your four conditions that you would impose on a targeted PRB, this targeted PRB, the question of the base, would those four conditions include a true-up mechanism of the kind you note or exclude a true-up mechanism? A. For the base in the first year? Q. Yes. A. That almost goes without saying that the starting point needs to be reliable and maybe for that reason I haven't explicitly highlighted it, but it's clear that the starting point of the PRB plan needs to be an Bauer 1130 re-ex (Warren) accurate reflection of the cost position. So, therefore, a hypothetical budget would not meet that purpose. I don't think it is necessary to true-up the base in years two and three, but the true-up of the initial starting point I think is essential if it is not possible to have actual values in place at that time of the -- starting of the plan. Q. In your discussion with Dr. Higgin you raised the -- I don't know if it was Dr. Higgin or you raised it, but it was the issue of significant -- I'll call it significant structural changes. You raised the possibility or said that if there are the prospect of significant structural changes it may not be appropriate to begin PRB now but wait until the structural changes have taken place. If the Board were to, following on that analysis, attempt to make a decision, what would the Board look at in determining whether or not it is appropriate now to begin it or to wait? ---[5:00 p.m.] A. If there are pending changes that will very likely will lead to significant restructuring of the company, the prudent decision would be to say, let's wait. Now, given the fact that the gas industry will probably not stop changing for the next decade or two decades, it's hard to say, I think it is necessary to make an assessment that the major changes will happen in the near future and then subsequent changes may be more Bauer 1131 re-ex (Warren) marginal or more gradual adjustments. If that's the case, I'm not deeply enough involved in the discussions in a number of other policy hearings that are currently conducted here to be able to assess that question. But if big significant changes will happen in the near future and then only small changes are expected in the following five years, perhaps the best strategy would be to envision a starting point that is somewhat postponed. If that's the case it may be a good opportunity also to envision a broader mechanism to begin with. Q. The final questions arise out of an exchange you had with Ms. Lea when Ms. Lea asked you by way of summary question whether or not if the changes you made is the proposal -- if the changes you had -- the conditions you had recommended, if they were imposed would the proposal be better than cost of service. In answering that question you introduced the notion of -- or the company if the cost of service is a transitory regime, is your word, to a comprehensive one, do you have any views, Dr. Bauer, as to whether in light of all of the evidence and discussion in this case what the duration, the ideal duration of the transitory term should be? Should it be, for example, three years as opposed to two years or one year? A. There is a significant tradeoff involved here. The tradeoff is between how long does it take for a Bauer 1132 re-ex (Warren) new incentive structure to take effect, and the common position here is that one year or even two years is too short really to activate incentives of new PRB plan, that many of the changes and adjustments that perhaps lead to efficient gains are longer term. Adjustments that will only happen if the planning horizon is stretched to three years. That needs to be traded off with the fact that it is more desirable to have a comprehensive PBR plan. My sense is, in order to get a PBR plan into workable shape and to allow the company and other stakeholders to adjust to it, to the new features, three years is probably an appropriate timeframe. If, for one reason or another, the transition period is seen as shorter than three years - two years or one year - then I think the argument in favour of a PBR plan, even if it is modified in the way I have outlined, really collapses and I think everybody is as well off in the current environment. Q. Okay. Thank you very much, Dr. Bauer. MR. WARREN: Those are my questions in re-examination. THE PRESIDING MEMBER: Thank you, Mr. Warren. Thank you very much, Dr. Bauer. ---Witness stands down THE PRESIDING MEMBER: I think that concludes our business. Supplementary Matters 1133 I would like to thank all counsel and the court reporters and Board Staff for all of their help. There are a couple of small things we need to put on the record. One is, I think you mentioned the changed schedule for argument for the previous case, and we haven't said anything about argument for this case. So I wonder if that's been discussed and if I can get any information on that? MR. CASS: It has been discussed, Madam Chair, or at least by way of a proposal being passed around to all counsel. We do, or did have a slight difference of opinion that, given the time of day, I was trying to avoid an argument about on the record. May I just speak with Ms. Allan for a moment, Madam Chair. I'm sorry. ---Off the record discussion. THE PRESIDING MEMBER: If you need some more time to discuss with other counsel, that would be satisfactory. MR. CASS: No, I think I have instructions, Madam Chair. First of all, may I go back to the unbundling case just to clarify. The Board will recall that the company had requested an extension to Thursday, February the 11th and the adjustments to the other dates in that schedule would result in parties -- intervenor arguments Monday, February Supplementary Matters 1134 the 22nd and company reply Wednesday, March 3rd. That was on the sheet that has been passed around, so I assume there's no difficulty with that. In terms of the argument for this particular case, the Applicant had proposed Monday, February 22nd for its argument-in-chief, the intervenors have requested Monday, March 8th for their arguments, and this was where we were having some difficulty, but I have instructions - in order to avoid an argument over this - and given the time of day and to allow us to move ahead - the company will live with the Monday, March 8th for intervenors' arguments and then actually that results in an adjustment to the reply argument date, which I haven't discussed with anyone else, but moving it back two business days would result in a reply argument date for this case of Thursday, March 18th. THE PRESIDING MEMBER: One moment, please. ---Off the record discussion. THE PRESIDING MEMBER: The Board is a little concerned with the length of the schedule but, given all of the other things that are going on, I think we'll agree to that and leave it at that. If there is any possibility of shortening it, I'm sure intervenors would have considered it by now. I assume that we will get affidavits for the very few witnesses that we didn't hear from on settled matters, if we haven't got them already? MR. CASS: Yes, Madam Chair. Supplementary Matters 1135 THE PRESIDING MEMBER: And are there any other matters? Ms. Lea, did you have anything to raise? MS. LEA: No, thank you, Madam Chair. THE PRESIDING MEMBER: Any other matters? Mr. Thompson? MR. THOMPSON: Yes. Can I just confirm that the outstanding undertaking responses will be faxed to interested parties? MR. CASS: That can certainly be done, Madam Chair. My last report, and I'm not even sure now how recent this was, was that 6.3 was being copied and the rest were here; leaving 4.2 and 6.4, which will be as finished quickly as possible and faxed. THE PRESIDING MEMBER: Thank you very much. And thank you again to everyone. We'll adjourn. ---Whereupon, the proceedings were concluded at 5:10 p.m. 1136 I N D E X o f P R O C E E D I N G S Page No. Preliminary Matters . . . . . . . . . . . . 970 On behalf of CAC, IGUA and OCAP: JOHN RANDOLPH NORSWORTHY, Ph.D.; Sworn. 972 Direct-Examination by Mr. Janigan 973 Cross-Examination by Ms. Lea 988 Cross-Examination by Mr. Cass 1003 Examination by Dr. Higgin 1048 Examination by Mr. Vlahos 1051 Examination by the Presiding Member 1059 JOHANNES M. BAUER, Ph.D.; Sworn. 1060 Direct-Examination by Mr. Warren 1060 Cross-Examination by Ms. Demarco 1086 Cross-Examination by Ms. Lea 1088 Cross-Examination by Mr. Cass 1101 Examination by Mr. Vlahos 1105 Examination by Dr. Higgin 1114 Examination by Mr. Vlahos 1120 Examination by the Presiding Member 1124 Re-examination by Mr. Warren 1129 Supplementary Matters . . . . . . . . . . 1133-1135 1137 L I S T O F E X H I B I T S No. Description Page No. I7.1 Document entitled "Benefits of Indexing". 1068 1138 L I S T O F U N D E R T A K I N G S No. Description Page No. H7.1 Dr. Bauer undertake to provide the tables underlying Exhibit No. I7.1. 1103 BD MC BV [ Copyright 1985]