Wednesday, 25 November 2009

Government outsourcing: the private sector can be good for you

At Emmanuelle Auriol and Pierre M.Picard have a new column, Government outsourcing: Public contracting with private monopoly. They open the column by saying,
Many countries, including Australia, New Zealand, the UK and the US, have chosen to outsource the investment and operation of non-competitive public services, such as water and waste management, public transports, mail services, information and communication technology services, and road infrastructures (Grout 2009). Outsourcing decisions imply the transfer of ownership and control of the facilities or services from public authorities to a private firm. However, economists do not usually recommend such transfers of natural monopolies. Outsourcing is no solution for the lack of competition that prevails in those sectors.
This last point does, therefore, raise the question, Why do governments make such contracts in non-competitive markets?
In the latest issue of the Economic Journal, we present a theoretical study on the costs and benefits of outsourcing in non-competitive contexts (Auriol and Picard 2009). Outsourcing contracts involve the transfer of control and cash-flow rights over a public service or infrastructure to a private firm in exchange for some investment. What distinguishes outsourcing from pure privatisation is that the government contracts with the private firm. We show that government outsourcing to the private sector is most desirable in advanced economies when it is used to deliver high-technology products or to cover low-profitability parts of public services. In very poor countries, outsourcing is beneficial when it leads to the creation of an infrastructure or service that would not otherwise exist.
So both developed and underdeveloped countries can benefit from government outsourcing. This result is based on an analysis which has at its base the government budget constraint captured by the shadow costs of public funds.
The shadow cost of public funds measures the social cost of the government's economic intervention; any transfer of public money to a firm implies either a decrease in the production of public goods, such as schooling and health care, or an increase in distortionary taxation. Shadow costs of public funds also reflect the macroeconomic constraints that are imposed on national governments' debts levels by supranational institutions (e.g. by the Maastricht treaty on EU member states, by the IMF on some developing countries) and the microeconomic constraints of government agencies that are unable to commit to long-term investment expenditures in their annual or pluri-annual budgets. In developing countries, low income levels and difficulties in implementing effective taxation programs are strong constraints on the government's budget, which leads to high shadow costs of public funds. Because our analysis focuses on the micro-economic outsourcing decision of a particular project, we take the shadow cost of public funds as given.
The Auriol and Picard study compares two regimes, a regulation regime and an outsourcing regimes.
In the regulation regime, a utilitarian government decides to set up a regulated firm run by a public manager. The government controls the investment and production decisions of the regulated firm and is therefore accountable for its profits and losses. Such a combination of control rights and accountability duties is typical of public ownership. The government designs incentive contracts to entice the firm’s manager, who has private information about the firm’s cost, to set the efficient level of production at some informational cost.
In the outsourcing regime, a private investor is invited to serve the market, possibly in exchange for a franchise fee. The private investor gains control and cash-flow rights on the outsourced activity. She controls the investment and production decisions and therefore is fully accountable for her profits and losses. As it is, the private firm is allowed to set the laissez-faire monopoly prices. Yet, because laissez faire is not necessarily optimal, the government can improve welfare by offering ex post contracts to the private firm; that is, once investment costs have been sunk and uncertainties have been solved. Ex post contracts are used by governments to entice the private firm to reduce its prices and increase sales. They are designed to fight the deadweight loss generated by monopoly pricing. However, to accept such ex post contracts, the private investor must at least obtain her laissez-faire profit, which raises her participation constraint to the scheme. The right panel of Figure 1 illustrates this point; the private firm’s profit under outsourcing (in blue) is everywhere larger than the private monopoly profit (in green).

What Auriol and Picard are able to show is that the optimal outsourcing contracts are more selective than the contracts under public management. By this they mean that low-cost private firms are offered ex post contracts that lead them to produce the regulated outcome, whereas high-cost private firms are not.
As a result the level of production under outsourcing is the maximum of the level of production under laissez-faire and regulation. The left panel of Figure 1 illustrates this point by showing the output levels under regulation (in red), laissez-faire (in green) and outsourcing (in blue).

This result is intuitive. At the contracting stage, the public manager of the regulated firm knows the cost parameter, whereas the government does not. She takes advantage of this information to obtain rents. To prevent managers of low-cost firms from inflating their cost reports, the government must reduce the output levels it asks to high-cost firms. Incentive issues can be so harsh that the output levels of high-cost firms become smaller than the output levels that they would achieve under laissez-faire. The left panel of Figure 1 illustrates this point by showing the firm with cost parameter β0 such that output is the same under regulation and laissez-faire. There is no point to offer an ex post contract to a private firm with this cost because its output level equals the government’s preferred output level. Ex post contracting and laissez-faire would yield the same consumer and producer surpluses. Consider next a firm with a cost larger than β0. If the government proposes an ex post contract to this firm, it is unable to get a surplus larger than under laissez-faire because incentive compatibility obliges it to distort its output downwards. Moreover, any transfer to this firm also increases the rents of all firms with lower costs. Because the government is harmed by both effects, it offers no ex post contract to firms with such large costs. The government thus has no obligation to subsidise the high-cost firms under outsourcing, and expected transfers to private firms are lower. In the right panel of Figure 1, the private firm’s profit under outsourcing (in blue) is smaller than the public manager’s rent under regulation (in red) for most cost parameters.

Outsourcing hence generates a positive fiscal effect because the government is able to terminate subsidies to those money-losing projects and possibly to collect a franchise fee from the private investor. Outsourcing also generates an economic surplus effect, as production can be higher under outsourcing than under a publicly managed firm, as illustrated by the left panel in Figure 1.

Comparing aggregate welfare in regulation and outsourcing scenarios, we show that the set of economic parameters supporting the outsourcing decision is far from negligible. Moreover, we find that outsourcing is more desirable for activities with stronger technological uncertainty or lower profitability and for governments with tougher financial constraints. This provides a useful grid for policy analysis of outsourcing decisions and public-private partnerships in practice.
Auriol and Picard use the pharmaceutical industry as an example of a high-technology industry that takes large investment risks and that receives, consistent with the optimal outsourcing scheme, ex post contractual arrangements from governments according to their effectiveness. These private pharmaceutical firms are able to choose their investments in R&D and the prices for their patented drugs. But their most effective drugs are subsidised by governments’ health insurance programs. Under such subsidies, consumption is higher than under laissez-faire, where no reimbursement is made.

Auriol and Picard also give examples of outsourcing in low-profit sectors.
Postal services are outsourced to grocery stores and to gas stations in rural areas of France, Sweden, and New Zealand. Our results suggest that such outsourcing policies are welfare-improving; the consumers get a service at extended business hours, the shop owners get additional revenue, and taxpayers shoulder lower subsidies. Similarly public transport is outsourced to local taxi companies in rural France, Switzerland, and Canada. The European Commission has subsidy programs that promote and finance such “taxibus” services. Taxi companies or drivers invest in their fleets and are free to operate their business. They nevertheless get subsidies to provide a public service that allow them to make higher profits than under laissez-faire. In low-density areas, this additional revenue is crucial for their economic survival. More importantly, this solution offers a better quality of service to the users at a lower public cost
Finally they discuss very poor countries. In such places
[...] outsourcing takes the extreme form of laissez-faire and is optimal for low-profitability segments. Sub-Saharan African water, electricity and transport services to the middle class and the poor are often offered by private enterprises without any subsidy. Because those countries have very tough budget constraints, even with the monopoly distortion, it is better than not to have a service at all.
  • Auriol, Emmanuelle and Pierre M. Picard (2009), “Government Outsourcing: Public Contracting with Private Monopoly”, Economic Journal, 119(540), 1464-1493.
  • Grout, Paul (2009), “Private delivery of public services,” Vox Talks,, 19 June.


Sally O'Brien said...

There is no such thing as a natural monopoly. All of the services you mention have or potentially have competition. ANY government sponsorship of ANY project takes away from the individuals' choice of where to spend their hard earned cash; to set their own priorities. This is especially onerous for low income people. That is why laissez faire is preferable to both total government funding AND government sponsorship and control. Both have a shadow cost on the rest of the economy. You say that otherwise certain services would not exist at all. Would that be because if they were funded voluntarily by individual consumers, those consumers would decide that they were too expensive?

Paul Walker said...

Sally. I would say there are few, but not zero, natural monopolies. Remember for a natural monopoly you need the cost function to be subadditive over the relevant demand range. A sufficient condition for this is for the demand curve to cut the average cost curve when average cost is still decreasing. It doesn't seem impossible for this condition to be met. The national electricity transmission network would be an example.

What is important, however, is that natural monopolies don't last forever. If either costs or demand changes, what is a natural monopoly today may not be one tomorrow.

Sally O'Brien said...

There are varied of alternatives for people who do not consider The national electricity transmission network to be a good deal. Even large enterprises such as factories can set up alternative generation for themselves especially if they have less burdonsome taxes and regulation. Any other examples?
Think before you speak.

Paul Walker said...

But I'm not talking about generation, rather transmission. Multiple generators is of course possible, but its not clear to me that, at least with current technology, more than one transmission network makes economic sense.

Sally O'Brien said...

I said there are alternatives to using the network - and there are. The electrical transmission network is not a monopoly - energy consumers have other options i.e. they do not have to use the network.

Paul Walker said...

The question of a natural monopoly is about the cost structure relative to the demand. Given that you have a transmission network, you will only have one. The alternatives to use of the network, given current technology, are not practical for the majority of people. Without a network how do you get power from the bottom of the South Island to Auckland. Given you need the network, you will need only one. The cost structure of the power network may be not subadditive, but I would doubt it.

Sally O'Brien said...

You don't necessarily need a network.

Paul Walker said...

With no network how do you get electricity from A to B?

Sally O'Brien said...

Oh use your imagination!
If I have solar panels and gas heating I don't need the network at home. If I own a factory and build a generator on site I don't need the network. If I run a shopping mall I can also build a generator on site. The network has to be efficient enough to compete with alternatives.

Paul Walker said...

Those alternative are as yet not cost efficient for most people. Getting power via the national grid is the most efficient way for most people. Given this you need a network but, and this is the important point, you need only one. The fact that the network has to be efficient enough to compete with alternatives doesn't mean it is not a natural monopoly. The nature of a national monopoly is that given you need something to be provided you only need one firm to do it.

I took this from the entry on Monopoly in "The Concise Encyclopedia of Economics" at the 'Library of Economics and Liberty':

"Natural Monopoly

David R. Henderson

The main kind of monopoly that is both persistent and not caused by the government is what economists call a “natural” monopoly. A natural monopoly comes about due to economies of scale-that is, due to unit costs that fall as a firm’s production increases. When economies of scale are extensive relative to the size of the market, one firm can produce the industry’s whole output at a lower unit cost than two or more firms could. The reason is that multiple firms cannot fully exploit these economies of scale. Many economists believe that the distribution of electric power (but not the production of it) is an example of a natural monopoly. The economies of scale exist because another firm that entered would need to duplicate existing power lines, whereas if only one firm existed, this duplication would not be necessary. And one firm that serves everyone would have a lower cost per customer than two or more firms."

Strictly speaking the defintion Henderson gives - unit costs that fall as a firm’s production increases - is sufficient but not necessary for a natural monpoly.

Over time if there are alternatives available the natural monopoly may be undermined, and this is what you see in many cases. The demand for the particular good or service could go to zero as the alternatives expand their market share. Or improved technology will alter the cost structure of production so that more firms can enter the market. Or demand could expand so that the efficient number of firms can increase.

But as technology and demand stands right now I would argue that the distribution of electric power is an example of a natural monopoly. Empirically Salvances and Tjotta (1998) estimate the cost function for electricity distribution in Norway and find that the distribution network is a natural monopoly.