The recent experience of some Western countries in reducing the shares of nationalized industries in economic activity and the transition from socialism in Eastern Europe have motivated many theoretical analyses of the privatization process. Both theoretical papers and case studies were presented at a conference organized jointly by CEPR and the Italian Macroeconomic Policy Group, which was held at the Innocenzo Gasparini Institute for Economic Research (IGIER), Milan, on 24-25 May. The conference was organized by Giorgio Basevi, a member of CEPR's Executive Committee, and Alberto Giovannini, a Research Fellow in the Centre's International Macroeconomics programme. Financial support for the conference was provided by Euromobiliare SpA and Prometeia.
Macroeconomics and Credibility
Carlo Favero (Queen Mary and Westfield College, London) presented the first paper of the conference, `Macroeconomic Effects of Selling State Enterprises', written jointly with John Driffill. He noted that the recent Commission of the Italian Treasury proposed a privatization programme with a view to reducing the public sector's borrowing requirement and hence solving the public debt problem. The recent theoretical debate on the macroeconomic implications of financing public debt has focused on bond finance and seigniorage, viewing the privatization of public assets as equivalent to the sale of government bonds with inverted streams of payments: if the issue is correctly priced with negligible transactions costs, privatization is equivalent to the sale of the enterprise's dividend stream.
Favero tested the validity of this assumption by estimating on UK data for 1979-89 a consumption function in which consumers maximize utility subject to a lifetime budget constraint. His results suggested that the public perceives privatized capital as net wealth but government bonds as neutral, even if the government views the two means of raising revenue as the same. Favero suggested that privatization may increase net wealth if it increases the productivity of privatized firms. Alternatively, privatization may reduce the government's incentives to `create inflation' by reducing its stocks of real assets and nominal outstanding debt, in which case a reduction in inflation-related inefficiencies may raise net wealth.
Guido Tabellini (Università di Cagliari, IGIER and CEPR) argued that privatization's macroeconomic effects are more likely to derive from future government behaviour than from the wealth effects considered in the `Ricardian equivalence' approach, which assumes that economic policy is exogenous. Mario Noera (Euromobiliare SpA) disputed the equivalence of a public enterprise's future stream of dividends with payments on outstanding government debt. Historically, many such enterprises have not provided governments with streams of income: indeed a major motivation of the Italian government's privatization programme is the desire to stop subsidizing public enterprises. Colin Mayer (City University Business School, London, and CEPR) noted that the UK government wrote off many public enterprises' debts in order to sell them, and their deliberate undervaluation entailed substantial wealth transfers to high-income groups, with significant implications for estimated consumption.Presenting a paper on `Successful Privatization Plans: Enhanced Credibility Through Timing and Pricing of Sales', written jointly with Serhat Guney, Enrico Perotti (Boston University) argued that state firms' lower profitability than comparable private firms does not justify their privatization, since they may be pursuing goals other than profit maximization although most such goals may be achieved through regulation rather than ownership. Sovereign governments are more vulnerable than private owners to lobbying from constituencies within the firm, and they cannot commit with respect to future regulatory policy. Since private owners can commit to reward efficient behaviour, privatization may provide a partial solution to this commitment problem: private property rights are constitutionally well protected and can credibly restrain but not eliminate governments' ability to interfere.
In Perotti's model of the privatization of a state enterprise, a `committed' government will seek to maximize the proceeds of the sale and then refrain from interference, while an `uncommitted' government will also take account of the potential gains from interference to redistribute the firm's surplus to a favoured constituency. Perotti showed that a committed government can improve its reputation with investors more rapidly by selling only a fraction of the shares in the first period. Where the perceived political risk is large, the government may also signal its commitment by underpricing the initial sale.
Alex Cukierman (Tel-Aviv University) noted that even an uncommitted government may refrain from interfering after the sale of the first tranche of shares in order not to reveal its intentions. Selling only a small proportion of shares in the first period will enhance a committed government's credibility, but the proportion sold will influence the behaviour of an uncommitted government and hence of investors who may expect such a government to be in office by the time of the second sale. Flavio Delbono (Università di Verona) maintained that Perotti's model overstressed the distinction between private and public sectors, since there are many kinds of legal status for firms across Europe. He added that if governments monitor public enterprises while regulatory agencies monitor private enterprises, there is an effective trade-off between the informational loss and the efficiency gain from privatization.
Optimal Sale Strategies
Thierry Verdier (DELTA, Paris) then presented a paper on `Privatisation in Eastern Europe: Irreversibility and Critical Mass Effects', written jointly with Gérard Roland. He noted that the current East European privatization programmes differ from those in the West because of their massive scale and because they start from a non-market environment. Such programmes may require a `critical mass' of assets in private ownership to get under way, and privatization may yield economies of scale if there are positive externalities associated with the size of the private sector. For example, a privatized firm may perform better in a market environment than in one with a substantial residual state sector; there may also be benefits in terms of non-rival inputs generated by R&D, as suggested by the literature on `new trade theory'; and a larger pool of managers may enhance shareholders' ability to distinguish the efficient from the inefficient. Such programmes may be jeopardized, however, by large-scale redistributions of income and wealth and by transitional unemployment.
Verdier presented a model of the transition from a socialist economy with identical workers and state firms, full employment and a fixed wage rate, in which high taxation is required to subsidize inefficient state enterprises, to a mixed economy, in which privatized firms maximize profits and the government maximizes the expected utility of the average citizen. Taxation on individual workers falls because fewer public firms require subsidies, but it rises because of the fall in the total number of workers employed. Private investors decide independently whether to buy shares. If they believe the government will abandon privatization at least in part in the face of political opposition, they will be less willing to invest, and a `low equilibrium' level of privatizations may result. Verdier argued that distributing shares free to all citizens eliminates the political risks associated with privatization by creating a `constituency in favour', but the externality associated with the size of the private sector remains.
Irena Grosfeld (DELTA) argued that the high costs of transitional unemployment are related not to privatization but rather to the reform of the state sector; the elimination of subsidies to state firms often precedes their privatization. Also, public understanding of the economic process in Eastern Europe is weak, there is no clear consensus that quick privatization is desirable, and workers will not necessarily perceive reductions to their personal taxation as the model assumes. Philippe Aghion (European Bank for Reconstruction and Development and CEPR) argued that revenues from asset sales and from minority shares may be substantial; and they may be better used to subsidize employment (as in the former GDR) rather than unemployment.
Thomas Chemmanur and Paolo Fulghieri (Columbia University) then presented their theoretical paper on `Privatization Under Incomplete Information: An Analysis of Equity Sale Strategies', in which they analysed the government's sale of a state-owned company to outside investors to maximize revenue when neither the government nor private agents know the firm's true value, but either may have private information. In their model of government, private investors and financial intermediaries, the government may sell at a fixed price or by auction; sell the entire equity at once or in two tranches; and use an underwriter or approach the equity market directly.
They found that selling shares by auction can be beneficial when the government has no private information, but this is available to private agents at low (or zero) cost. When the government has favourable private information about the firm's value, however, it may best convey this information credibly to potential investors by selling its shares in tranches at fixed prices and deliberately underpricing the initial offer. A successful initial sale will enable the government to make a gain in the second offer, once additional information concerning the performance of the privatized firm has become available. The authors also noted that while the government has greater risk-taking ability than any group of underwriters, so investment bankers may seem unnecessary for privatizations, financial intermediaries may nevertheless play a useful role if they have significant expertise in valuing equity and sufficient reputation to convey it credibly to investors.
Marco Pagano (Università di Napoli and CEPR) suggested extending the model to consider the role of public enterprises' managers, who will typically hold better private information than the government and have no incentive to reveal it to the government at the time of sale. If it is favourable, they may leak it to friendly investors and split the proceeds; if it is not, they would risk being sacked by revealing it. Pagano also questioned the assumption that the firm's true value is known after the sale, since this will still depend on the government's behaviour concerning tariffs and regulations. Alessandro Penati (Università Bocconi, Milano) pointed out that the government's private information here will be more important than that about the privatized firm.
Case Studies
Matthew Bishop and John Kay (London Business School) presented their joint paper on `The Impact of Privatisation on the Performance of the UK Public Sector'. They first noted that privatization in the UK gained a momentum of its own in the 1980s, with four stated objectives: to raise government revenue; to reduce the power of public sector unions; to promote wider share ownership; and to increase efficiency. Focusing on the latter, the authors found that firms that had been privatized were more profitable than those that had not; but profitability only indicates efficiency for firms with `private sector' objectives that operate in competitive markets, such as British Airways and British Steel. It is unsuitable for utilities or other industries that were nationalized on account of wider perhaps ill-defined social needs.
Measuring firms' efficiency instead in terms of total factor productivity (TFP), they found that TFP growth has been much greater since 1983 than during 1979-83, but this growth appears wholly unrelated to privatization. Gains in British Coal, British Rail and British Steel (which was not privatized until December 1988) have bettered or at least matched those of the flagship privatized enterprise, British Telecom; while British Gas one of the earliest utilities to be privatized has been the poorest performer since 1983. Indeed, the most spectacular improvements suggest causality running from improved performance to privatization. Bishop and Kay argued nevertheless that these TFP gains could not have been achieved without the UK government's commitment to impose commercial standards on the nationalized industries, and that privatization played a critical role in signalling that commitment.
Vittorio Grilli (Birkbeck College, London, and CEPR) questioned whether 1979-90 is a long enough period for the effects of privatization on performance to become established, since on average the `1979 public sector' has only been privately owned since 1986. He also suggested comparing the performance of industries now under mass private ownership with that of the few firms privatized through private sales, whose change of ownership may influence management decisions more quickly. Colin Mayer stressed the need to distinguish welfare gains from wealth transfers and expressed concern that the authors said little about consumer prices, wage rates or quality of services. He suggested that most of the improvement in quality of service was attributable not to privatization but rather to improved regulation. Regulators now have better incentives and clearer limitations on their role, although their separation from the former nationalized industries' management entails costs in so far as resources must now be devoted to `regulatory games'.
Guillermo de la Dehesa (Banco Pastor, Madrid, and CEPR) presented a paper on `European Privatization: The Case of Spain', in which he noted that when Spain embarked on its privatization programme it already had the smallest public sector in the European Community. This programme has been characterized more by `learning-by-doing' than by any model, theory or plan. It was not motivated by any deliberate policy to reduce the public sector, as in the UK or in France: indeed a number of renationalizations have taken place. The government aimed to use privatization to overcome a variety of strategic, technological and budgetary constraints by rationalizing the structures of companies whose nationalizations had been forced by political and regional problems during Spain's isolation under Franco. It also sold a number of large corporations that had no future in state hands because their technology, scale of production or distribution capacities were inadequate to compete in international markets. These were mainly sold to foreign buyers, either by auction or by direct negotiation with a few selected candidates, with a view to securing their long-term commercial viability and with remarkably little concern for `nationalistic' interests.
Leonardo Felli (Boston College) argued that private owners will have a greater stake than government in assets and are more likely to invest in privatized firms in a human-capital-specific manner. Any government that has nationalized a large firm in order to prevent its financial distress can have little motivation to privatize, however, since it has destroyed the credibility of the bankruptcy constraint. Xavier Freixas (Université de Toulouse) suggested that under the `pragmatic' approach to privatization the `optimal' size of the public sector may change in response to exogenous factors, such as the business cycle and Spain's integration into the European Community. Maintaining such an `optimum' may require a programme of renationalizations in ten years' time.
Presenting the final paper of the conference, on `Privatisation in France', Robert Delorme (Université de Paris Nord and CEPREMAP, Paris) noted that while cost efficiency is commonly an important factor in models of the rationale for privatization, it has never been in the forefront of French industrial policy. This has stressed instead the importance of competitiveness both at the firm level whether private or public and at the international level. Studies of French privatization that focus on the changes of ownership under the right-wing government of 1986-8 therefore miss many significant longer-term changes to France's traditionally interventionist industrial policy. These appear in changes to the management methods and objectives that public sector firms are required to follow rather than in changes in ownership.
The 1986-8 programme of `denationalization' was nevertheless significant as a political response to the 1981-2 programme. It aimed to privatize two-thirds of public enterprises and 40% of public sector employment within five years, and almost 50% of the programme took effect within the 16 months of the right-wing government's `cohabitation' with the Mitterrand Presidency. Many of the programme's economic objectives were addressed, however, by policy measures already in place. For example, public enterprises' finance had been drawn from sources wider than the public sector budget since 1983, and their use as instruments of macroeconomic stabilization was already vanishing in 1982-3. The re-election of a socialist government has not led to a renewed programme of nationalizations, but rather to a pragmatic policy of neither privatization nor nationalization.
Patrick Rey (ENSAE, Paris) noted difficulties in measuring changes in enterprises' `priority management methods' as discussed in Delorme's paper and noted the specificity of French industrial organization: in particular, the Chief Executive Officers of France's 13 leading non-financial companies all began their careers in the public sector. Richard Portes (CEPR and Birkbeck College, London) noted that there has been no effective change in the structure of French firms and questioned how increased competitive pressures could therefore be exercised. If ownership does not provide an incentive to competitiveness, it is unclear how to measure whether or not the 1986-8 ownership changes achieve their expected results.