Alternative approaches to regulation: an economic analysis of lighthanded regulation

1
Alternative approaches to regulationan economic
analysis of light-handed regulation

• Simon Cowan
• University of Oxford
• ACCC Regulatory Conference
• 26-27 July 2007

2
What is light-handed regulation?

• Formalized in New Zealand for telecoms regulation
• General competition law
• Disclosure requirements
• Threat of subsequent specific regulation
• For Telecom New Zealand some additional
  constraints on pricing
• In other versions there is the possibility of
  binding arbitration in the event of disagreement

3
Some theory

• Three reasons for specific regulation
• To avoid deadweight losses
• To promote distributional objectives
• To encourage relationship-specific investments
• Asymmetric information and incentives
• No regulation can be better than standard types
  of regulation
• Other theoretical examples where regulation gives
  poor outcomes input choices, relative prices
• The threat of regulation

4
Why regulate?

• Avoidance of deadweight loss ( unrealized gains
  from trade)
• But regulation itself has costs
• It requires specialists (who may be particularly
  costly in developing countries)
• There are probably economies of scale in
  regulation, so it is more worthwhile in larger
  markets
• There may also be economies of scope gt
  multi-sector regulation

5
Bargaining instead?

• The Chicago/Coase question
• Why dont the firm and its customers bargain to
  realize all the gains from trade, and thus
  eliminate the deadweight loss?
• Standard counter-argument
• Customers are typically too small, and too
  numerous, for bargaining to be effective
• Maybe we can rely on bargaining when the
  customers are themselves large firms
• Bargaining and distribution regulation to
  allocate the gains from trade?

6
Bargaining and access

• A network firm bargains with access seekers, but
  no-one bargains with final customers
• The firms have an incentive to maximize joint
  profits, so bargaining is efficient for the firms
  but not for society
• We want firm-to-firm bargaining to be less than
  fully efficient
• The possibility of secret deals (private
  renegotiation) reduces the available profit
• asymmetries of information on both sides reduce
  bargaining efficiency
• King and Maddock (1999) threat of binding
  regulation with lower profits, and constraints on
  the bargaining process, can dissipate actual
  profits in equilibrium
• Overall the argument for bargaining is rather
  delicate either we want it to be perfect for all
  players, or if it is firm-to-firm we want it to
  be less than fully efficient so that joint
  profits are not maximized

7
Regulation as a substitute for long-term contracts

• There are relationship-specific assets, often on
  both sides, and long-term contracts cannot be
  used
• A new electricity connection involves
• (i) wiring in the customers premises
• (ii) direct connection to the network
• (iii) network reinforcement (possibly)
• Typically these costs are shared, with the
  customer paying (i) and (ii) which may be a
  uniform price and the firm paying for (iii)
• The legislation behind specific regulation, the
  agency tasked with implementing it and the
  judicial framework all provide some stability for
  both sides
• Such features may be lacking in light-handed
  regimes

8
A model of incentives and efficiency

• Demand is downward-sloping, so deadweight losses
  can occur
• Marginal cost is subject to shocks, which are
  private information, and can be reduced by effort
  (which is costly and private)
• Regulator chooses between three forms of
  regulation rate-of-return, price-cap,
  light-handed

9
The set-up

• Demand is linear in the price
• Marginal cost (MC) Cost shock Effort
• Social welfare Consumer surplus Profits
• Full information benchmark has price equal
  average cost

10
Three types of regulation

• Rate of return (ROR) constant margin
• Price cap (PCR) constant price
• Light-handed regulation (LHR) no regulation
• The firm is free to choose its price and effort
  to maximize profits for each cost shock
• Profit must be non-negative whatever the cost
  shock
• The aim is to compare expected welfare under the
  three forms of regulation

11
Rate-of-return regulation

• Price Fixed margin MC
• Some effort is valuable for the firm if the
  margin is positive it reduces MC, which reduces
  the price, which increases demand, and the firm
  then earns a fixed margin on the extra sales
• Two offsetting effects of the margin on the
  price
• Direct effect higher margin raises the price
• Indirect effect higher margin leads to more
  effort, which cust MC and thus the price
• These two exactly offset each other Price Cost
  shock
• Regulator can set the margin optimally to
  maximize expected profits

12
Light-handed regulation

• The firm chooses price and effort to maximize
  profit in each state
• Now more effort has an effect on the price
• The firm exploits its market power
• But effort cuts MC, which reduces the monopoly
  price
• The overall effect is that
• The price with LHR is the same as with ROR
• LHR dominates ROR in welfare terms, and not just
  on average
• Why? The firm responds to the cost shock
  optimally under LHR, but will not change its
  effort level under ROR.

13
Price-cap regulation

• Regulator fixes the price so that average cost in
  the worst state is just covered
• PCR is fine if there is no asymmetry of
  information (the cost shock is known and
  constant)
• As the cost shock becomes more variable it
  becomes more likely that rate-of-return
  regulation is better than price-cap, and thus
  that light-handed is better than the price cap

14
The lesson

• In this model, with restricted regulatory
  options, light-handed regulation is better than
  rate-of-return, and can be better than price-cap
• Light-handed regulation allows the firm to
  respond to cost shocks
• With demand shocks such flexibility will usually
  be damaging
• Well-intentioned regulation can unintended
  consequences
• But we can always devise even better regulation
• We could allow something in between full or zero
  passthrough
• Better than either light-handed or price-cap
  regulation is a combination price ? price-cap

15
Specific regulation can be worse than nothing

• Averch-Johnson (1962) effect
• Rate-of-return regulation
• Allowed rate exceeds the cost of capital
• Costs are not minimized
• Avoided if regulation fixes the price and reviews
  are stochastic
• Price caps and relative prices (Cowan, 1997,
  1998)
• A tight cap on average revenue can induce the
  firm to set excessive prices in some markets to
  alter the weights in the index and allow prices
  in other markets to rise
• This can also occur with a regulated price index
  with fixed weights here the problem is
  excessively low prices in some markets
• In both cases regulation is more likely to be
  worse than no regulation when the level of the
  price cap is low regulation is tough
• Avoided if price index is of Laspyeres-type
  (weights proportional to quantities and updated
  each period) the tariff basket

16
The threat of regulation

• Same as a firm choosing its price facing a
  possible fine for anti-trust abuse based on
  excess profits
• If the probability of regulation (or of
  conviction) does not rise with the price the
  threat of regulation has no effect on pricing
• The rate of increase of the probability as the
  price rises, as well as the level of the
  probability, matter for deterrence
• The threat of regulation can achieve some of the
  outcome of direct regulation without the
  associated costs
• But this doesnt allow for investment decisions

17
New Zealand telecoms

• Formalized light-handed regulation
• Problems over access negotiations between Telecom
  and Clear
• Recent move to specific regulation
• Interconnection in other countries
• Split of ATT in 1984 partly because of
  anti-competitive behaviour by ATT towards rivals
• Pro-competitive determination of access prices
  for Mercury and British Telecom in 1985 in the
  UK, and subsequent direct regulation of access
  terms
• Local-loop unbundling and broadband access the
  modern version of an old story of interconnection
  controversy
• Verdict interconnection difficulties were
  predictable, maybe the possibility of binding
  arbitration should have been introduced

18
New Zealand electricity

• Bertram and Twaddle (Journal of Regulatory
  Economics, 2005) do a careful accounting exercise
  to determine the excess profits earned by
  electricity distribution networks over 1995-2002
• Average variable costs fell from 1.85 to 1.25
  cents/kWh in real terms
• Output rose over this period by 18
• The physical asset stock was largely unchanged
• Average cost pricing would have normally caused
  prices to fall
• But average revenue rose from 3.49 to 3.88
  cents/kWh
• The asset base rose by 3.6 billion, of which
  2.9 billion was due to unilaterally declared
  asset revaluations
• The authors calculate that there was a 200
  million over-charge each year, compared to what
  would have happened under a standard
  rate-of-return regime

19
Some problems

• The study uses average revenue, but actual
  tariffs are multi-part, so the rent transfers
  identified need not imply allocative inefficiency
• The study assumes (a) that the light-touch regime
  could have been replaced by perfect
  rate-of-return regulation that fully extracted
  rents and (b) that there would have been no
  resulting change in firm behaviour (no incentive
  effects)

20
Australian airports

• Forsyth (2006)
• Initial tough CPI X price caps for the airports
  once corporatized and privatized (real price
  reductions of 20-25)
• Tariff-basket used
• Losses incurred even before the shocks of
  September 2001
• Replaced by light-touch regulation in 2001/2
• After a review in 2007 the regime is to continue
  until 2012
• Deadweight losses probably unimportant, airport
  objectives not only profit-maximization
• A success story for light-handed regulation?

21
UK gas

• British Gas was privatized in 1986
• De facto light-handed regulation but with a
  specific regulator
• CPI 2 for transportation element of retail
  prices full pass-through of gas purchase costs
• No regulation of prices to larger customers
• Weak regulation of access regulator to determine
  price if no agreement
• No competitive access for five years
• Subsequent heavy intervention
• Four investigations by the competition
  authorities (1988, 1991, 1993, 1997) leading to
  vertical separation
• Tighter retail and access price regulation
• Initial regulatory framework was unsustainable

22
Conclusions

• Light-handed regulation has some attractions in
  theory
• Saves on resource costs, the threat of regulation
  can keep prices low
• In practice the record is mixed
• New Zealand has moved away in telecoms and
  electricity
• But the Australian airports example appears to be
  successful