Dr. Alejandro Díaz-Bautista, Professor of Economics and Industrial Organization. Regulation and the Averch Johnson Model.

1. Industrial Organization Dr. Alejandro Díaz-Bautista Professor of Economics and Industrial Organization [email_address] Regulation and Averch Johnson Model Dr. Díaz-Bautista received his Ph.D. in Economics from the University of California Irvine (UCI). He also earned his master's degree in economics at UCI. He was also educated at UCSD and ITAM in Mexico City where he earned his Bachelor’s degree in Economics. His career has involved academics, government service and consulting for private firms.

2. INTRODUCTION Regulation limits the actions of economic agents e.g price controls, standards/qualifications, licences, sale restrictions. Focus on regulation of ‘natural’ monopolies usually ‘essential’ industries - telecoms, water, gas, electricity, rail. Usually publicly-owned, statutory monopolies In last 20 years, privatisation of state-owned monopolies and liberalisation of industries. In many industries, ‘natural’ monopoly in some sector  need for regulation remains (networks). but other sectors can be competitive, e.g. phone calls, gas supply, rail services.

3. Why regulate? Focus on 2 types of market failure (i) monopoly power: ability to charge p > MC p > p c  dwl > 0 may be excess capacity  firms not at min. AC price regulation   p c (1st best) also the aim of competition policy. (ii) asymmetric information. one party has more info. than another. e.g. firm has more info re. costs, demand, technology. rational choice not possible as true quality and/or price not known Focus on situation where ‘natural’ monopoly exists, and there is asymmetric info.

5. necessary that average total cost is decreasing over large range of desired output (ES, IRS) IRS natural monopoly if hetero. goods (monopolistic competition) competition  duplication of costs and inability to achieve large ES Nat.Mon. requires that prices and average costs are lower with only one firm serving whole market i.e. sub-additivity of costs (Braeutigam p.1295) Suppose there are n products and k firms: Each firm can produce all n products. Let be the amount of r produced by firm i, (r = 1,…..,n and i = 1,….,k). Let be firm i’s output vector.

6. A cost function C( y ) is strictly subadditive at y if, for any and all quantities of outputs y 1 ,…., y k , y j  y , j = 1,…,k, such that , we have ‘ An industry is a natural monopoly if, over the entire relevant range of outputs, the firm’s cost function is subadditive’ - Baumol, Panzer and Willig (1982). subadditivity

7. IRS over all output levels Profit max pricing inefficient as firm sets p m > MC 1st best  p = MC < AC   < 0  no prod. Can govt. improve the situation? 4 choices. 1. set = MC but subsidise losses if fixed/sunk costs are positive 2nd best sol. as subsidy from distortionary tax. 2. Public ownership: run at a loss . 2 nd best as resources could be used elsewhere. 3. Ramsey pricing: set  AC    0  Q > 0 2nd best as > MC. price set too low  may prevent competition and/or innovation in long run. p is a uniform tariff. Can improve situation by imposing non-linear tariffs, e.g. bill-pay v ready-to-go on mobile phones

8. Introduce competition May be desirable if markets are ‘contestable’ - requires no sunk costs, free entry/exit p  AC to deter entry (still 2nd best) If competition within market is impossible, govt. can still ensure 2nd best solution w/o fixing prices. Transfer firm surplus to consumers by auctioning production rights or offering franchises i.e. competition for the market (Demsetz comp.) If p* is lowest break-even price, firms bid p  p*. Large no of firms imply winning bid will be p  p* = AC > MC (lowest cost firm wins). Need to enforce contract may be costly - monitoring costs, unspecified contingencies

9. TYPES OF REGULATORY REFORM (i) Structural change: can maintain or alter industry structure e.g. privatise, break up monopoly horizontally and/or vertically . (ii) liberalisation - open markets to competition. e.g. allow entry to telecommunications industry . (iii) conduct regulation - control market power by restricting firms actions/strategies e.g. fixed access and product prices (telecoms), quality regulations (water, electricity supply). to be effective, need info. re. future costs and demand initially, focus on (iii)

10. In US, rate-of-return regulation. UK has tended to use price-cap regulation. Rate of return regulation : Utility calculates costs and regulator sets r.o.r on capital regulator then sets prices to ensure the firm raises enough revenue No incentive to min. costs, price rise if r < r*. May be solved by imposing a regulatory lag so firms keep cost savings. r* too high (low)  firms may over (under) invest in capital to make higher profits (reduce quality). Problem of measuring capital - historical or replacement costs?

11. Price-cap regulation : Regulator collects firm info. to determine revenue requirement, then sets form of price control for y years Price regulation faces trade-off between : (a) giving firms incentives to minimise costs. (b) passing cost reductions on to consumers. Optimal price regulation takes form of price-cap regulation known as ‘ RPI-X’ or ‘CPI-X’ thought to give better incentives for cost reduction. States by how much nominal prices can rise in different periods. RPI & CPI are measures of inflation. X denotes expected efficiency/productivity gains (real price changes) - usually firm-specific.

12. If RPI < X, nominal prices must fall Expected cost reductions are passed to consumers, while firm keeps reductions > X (‘beats the cap’). But, how does regulator decide on X? depends on costs, future demand and productivity. X too low (high), may  high profits (loss). If X re-set mid-term to offset high profits, may reduce future incentives to cut costs. For multi-product firms, regulation may be imposed on average prices, e.g. Eircom. When privatising, X likely to be set in favour of consumers. After that, due to asymmetric info., lobbying etc., may favour producers.

13. RPI-X regulation more suited to industries where techno. is changing rapidly X can be adjusted to changes in cost and market conditions RPI - X used in telecoms and electricity industries. RPI - X + Y used in gas industry, where Y is % increased gas costs that can be passed on to consumers, e.g. fuel. In water industry, have used RPI + K, where K represents cost required to maintain quality standards.

14. Problems of Regulation setting 1st or 2nd best price implicitly assumes that regulator and firm have same info. Firm likely to have better info. re. demand & costs May be costly for regulator to acquire this info. Firm may over-state costs in order to benefit from a higher access or final product price. Also, problem of regulatory capture . Regulator acts in interests of firms rather than society as a whole. Likely when regulator depends on firm for info. may be due to lobbying, corruption. regulator may be former employee/lobbyist of regulated firm or promised position afterwards

15. If asymmetric info., principal-agent theory can describe relationship between regulator and firm. The regulator must design a regulatory policy that induces the firm to act as the regulator wishes it to. Policy must satisfy the firm’s (a) Participation constraint – firm’s payoff exceeds some reservation level (b) Incentive compatibility constraint – firm will act ‘truthfully’, e.g. will not mis-state costs, choose sub-optimal effort.

16. The Averch-Johnson Effect H. Averch and L. Johnson (1962) &quot;The Behavior of the Firm Under Regulatory Constraint,&quot; American Economic Review , December 1962 . Averch and Johnson developed a model to illustrate that public regulation creates an incentive for the firm to over-invest in tangible assets . Since the &quot;allowed profit&quot; is based on the rate base (RB), the firm has an incentive to augment its capital stock.

17. Over-investment (or over-capitalization) has obvious implications for rates paid by consumers and also for the efficiency of resource allocation.

18. Choose quantities of capital and labor to maximize the following profit (  ) function:  is profit R is the revenue function K is the quantity of capital L is the quantity of labor w is the wage rate r is the cost of capital s is the allowed rate of return The Averch Johnson model [1] subject to [2]

19. Averch Johnson assumption: s > r This would seem a logical assumption--why would the firm take positions in tangible capital goods (like nuclear plants) if r > s ? Meaning the allowed rate of return on capital (expressed in dollars per unit of capital per time period) exceeds the cost of capital (also dollars per unit of capital is the same time interval).

20. Maximizing [1] subject to [2] using the Lagrangean method yields the following first order condition: MP k is the marginal product of capital MP L is the marginal product of labor  is the Langrangean multiplier (a constant). [3] [4] Note that:

23. Q = 100 Q = 300 Q = 200 Labor (units) Capital (units) 0 Isoquants    is a labor intensive technique  is a capital intensive technique

24. Labor (units) Capital (units) 0 20 25 100 Intercept = C/w = $1000/10 Intercept = C/r = $1000/50 Slope = -w/r r = 40 Let C = wK + rK, where: C = $1,000 w = $10 r = $50

25. The Averch Johnson Effect 0 Capital (units) Labor (units) M M ’ isoquant E A Slope = -r/w Slope = -(r -  )/w E is an efficient point A is the Averch Johnson point N N ’

28. Revenue Requirement Model RR is the “revenue requirement” RB is the rate base—an estimate of the value of owners’ investment in the regulated firm. r is the “allowed rate of return” to owners’ investment. E is expenses (fuel, wages, etc.)

29. Electric deregulation is really about the vertical separation of the three stages of production—or the creation of a multi-market industry .

33. First year under new regime Average Retail Electricity Prices in California (cents per kilowatt hour) Source: California Energy Commission 13.92 2003 13.41 2002 13.30 2001 10.42 2000 10.11 1999 10.09 1998 10.06 1997 9.98 1996 9.94 1995 10.23 1994

70. Industrial Organization Dr. Alejandro Díaz-Bautista Professor of Economics and Industrial Organization [email_address] Regulation and Averch Johnson Model Dr. Díaz-Bautista received his Ph.D. in Economics from the University of California Irvine (UCI). He also earned his master's degree in economics at UCI. He was also educated at UCSD and ITAM in Mexico City where he earned his Bachelor’s degree in Economics. His career has involved academics, government service and consulting for private firms.