Empirical evidence on privatisation

On RogerKerr.wordpress.com it is noted that

[...] the unequivocal findings of economic research are that on average and over time, privately owned businesses outperform publicly owned ones (see here for relevant references). What matters for policy is this general result. Government should not bet against the odds with taxpayers’ money.

In addition to the Phil Barry reference Kerr gives the empirical evidence on privatisation is discussed in William L. Megginson's book, The Financial Economics of Privatisation, Oxford University Press, 2005.

The 87 studies from nontransition economies discussed in this chapter offer at least limited support for the proposition that privatization is associated with improvements in the operating and financial performance of divested firms. Most of these studies offer strong support for this proposition, and only a handful document outright performance declines after privatization. Almost all studies that examine post-privatization changes in output, efficiency, profitability, capital investment spending, and leverage document significant increases in the first four measures and significant declines in leverage.

The studies examined here are far less unanimous regarding the impact of privatization on employment levels in privatized firms. All governments fear that privatization will cause former SOEs to shed workers, and the key question in virtually every case is whether the divested firm's sales will increase enough after privatization to offset the dramatically higher levels of per-worker productivity. Three studies document significant increases in employment [Galal, Jones, Tandon, and Vogelsang (1992); Megginson, Nash, and van Randenborgh (1994); and Boubakri and Cosset (1998)], but most of the remaining studies document significant-sometimes massive- employment declines. These conflicting results could be due to differences in methodology, sample size and make-up, or omitted factors.

However, it is more likely that the studies reflect real differences in post-privatization employment changes between countries and between industries. In other words, there is no "standard" outcome regarding employment changes.

Perhaps the safest conclusion we can assert is that privatization does not automatically mean employment reductions in divested firms, though this will likely occur unless sales can increase fast enough after divestiture to offset very large productivity gains. Since the empirical studies discussed in this chapter generally document performance improvements after privatization, a natural follow-up question is to ask why performance improves. For utilities, the need to introduce competition and an effective regulatory regime emerges as key, but there is no "silver bullet" answer for what makes privatization successful for firms in competitive industries. As we will discuss in the next chapter, a key determinant of performance improvement in transition economies is bringing in new managers after privatization. No study explicitly documents systematic evidence of this occurring in nontransition economies, but Wolfram (1998) and Cragg and Dyck (1999a,b) show that the compensation and pay-performance sensitivity of managers of privatized U.K. firms increases significantly after divestment. Studies that explicitly address the sources of post-privatization performance improvement using data from multiple nontransition economies tend to find stronger efficiency gains for firms in developing countries, in regulated industries, in firms that restructure operations after privatization, and in countries providing greater amounts of shareholder protection.

Another overview of the empirical literature is Sunita Kikeri and John Nellis's An Assessment of Privatization, "The World Bank Research Observer", vol. 19, no. 1 (Spring 2004)

This article takes stock of the empirical evidence and shows that in competitive sectors privatization has been a resounding success in improving firm performance. In infrastructure sectors, privatization improves welfare, a broader and crucial objective, when it is accompanied by proper policy and regulatory frameworks.

Mary M. Shirley and Patrick Walsh write in Public versus Private Ownership: The Current State of the Debate, Working Paper, The World Bank,

Our review found greater ambiguity about ownership in theory than in the empirical literature. In the debate over the effects of competition, theory suggests that ownership may matter and if so, that private firms will outperform SOEs. The empirical studies squarely favor private ownership in competitive markets. Theory’s ambiguity about ownership in monopoly markets seems better justified, since the empirical literature is also less conclusive about the effects of ownership in such markets. Theories that assume a welfare maximizing government suggest that SOEs can correct market failures. In contrast, public choice theories are skeptical of the benevolent government model. Corporate governance theories suggest that even well intentioned governments may not be able to assure that SOE managers do their bidding. The empirical literature favors those skeptical of SOEs as a tool to address market failures. In studies of industrialized countries, where we might expect more developed political markets to motivate greater government concern with welfare maximization or better information and incentives to overcome corporate governance problems, private firms still have an advantage. The private advantage is more pronounced in developing countries, where market failures are more likely.

As to the empirical evidence for New Zealand let me deal with one obvious recent and controversial example: Kiwirail.

In the July 2009 issue of Competition and Regulation Times put out by the New Zealand Institute for the Study of Competition and Regulation (ISCR) the question is asked, Kiwirail: strategic asset or strategic blunder? The article summaries an ISCR research paper "The history and future of rail in New Zealand" by Dave Heatley.

We seem to be nearer the blunder end of the scale than the asset end. The Times article and the research paper argue along similar lines. Heatley opens his Times article by noting that back in 1999 one of the first projects undertaken by the ISCR was a study of the long-term economic performance of New Zealand railways.

Public rail ownership was characterised by declining performance, beginning in the 1920s and culminating in a very poor prognosis in the 1990s. There were signs that since 1993, privatisation had led to improved productivity and profitability; however, the business was still far from achieving financial sustainability. The ISCR report predicted that private-sector ownership would result in better incentives for productivity-enhancing decision making, but in the long run it was unlikely that in its current form the business would be able to generate returns sufficient to cover the costs of the very large sums of capital employed. Given these facts, a rational private owner would likely rationalise services and reduce the scale of the network to the point where it constituted a sustainable long-run business. Revenues freed up from repeated cycles of historic government-funded capital injections and operating subsidies could then be applied to more productive uses, to the wider benefit of the New Zealand economy.

Given that rail is again in the hands of the government it is timely to re-examine the assumption that government ownership will result in superior long-term outcomes for the long suffering taxpayer owners. Heatley writes,

The 2009 analysis reveals little evidence to suggest that overall the economic outlook for rail has improved since 1999. Despite gains in operational productivity, rail's share of the land freight task has declined over the period examined. Profitability has remained poor, suggesting an ongoing lack of competitiveness vis-a-vis other freight modes.

and continues

Rail networks offer benefits from economies of density (increasing use of existing tracks), but not necessarily from economies of size (increasing size of the network).' In a rail network with uneven patterns of use, such as New Zealand's, the economics of density means that the closure of lightly used lines will, in general, improve the overall economic performance of the network.

Importantly Heatley notes that

It proved difficult for private owners to rationalise the size of the network efficiently, due to poorly aligned incentives and political intervention in operational decisions such as exiting from the provision of certain long-distance passenger services.

The retention of land ownership by the Crown at the time of privatisation muted private incentives to rationalise the network as the private operator was unable to access the potential land-sale benefits from closing unprofitable lines. Private-sector owners have been incentivised to persevere with a strategy (originating under public ownership) of retaining otherwise uneconomic lines for their current income-generating potential, but refraining from investing in replacement infrastructure such as sleepers, tracks and bridges.

A return to integrated land, infrastructure and operational ownership resolves the incentive misalignment, enabling its new owners to rationalise network infrastructure efficiently. Yet perversely, extensive recapitalisation has followed re-nationalisation. The government has invested $2.9 billion in rail since 2002, and has committed a further $0.9 billion through to 2013. It is unlikely that the government will earn a reasonable financial return on this investment, as the strong incentives of private owners for ongoing productivity improvements will likely be muted under government ownership, and the scope for political intervention in strategic and operational activities has increased.

The consequences of political intervention are evidenced in the targets set for a modal shift from road to rail freight in the New Zealand Transport Strategy. Any increases in rail freight's share must ultimately come from substitution at the margins away from competing transport modes. Extensive competition from both road and sea freight restrains the ability of rail to set prices. Rail exhibits few apparent cost advantages, even with subsidies from the written-off opportunity cost of capital. So modal shift can only be driven by increasing the level of subsidies in order to lower prices artificially and therefore induce movement of marginal freight away from more efficient road and sea freight. Such shifts will be to the detriment of the overall economic performance of the transport sector and the wider New Zealand economy.

There is little evidence that the real costs of the current government ownership and investment strategy have been adequately assessed in terms of foregone benefits in other taxpayer-funded areas, such as health and education.

After this, an obvious question to ask is, Is there light at the end of the tunnel? Heatley comments,

The 2009 analysis confirms that the issues identified in 1999 still remain, and are unlikely to be addressed by recent changes in governance, ownership and policy direction. Yet rail still remains a viable transport medium for those segments to which it is intrinsically well-suited - long-haul carriage of heavy, bulky freight (coal, logs, manufactured goods, etc.) and high volume urban commuter services. The challenge for rail's new owners is to find a viable subset of the current rail network. Given current and projected freight and passenger types and volumes, it appears a viable subset exists at around 1500-2000 kilometres in length - less than half the present size. Line closures and land sales could fund upgrading of the core network to 21st-century standards.

So, rail makes sense for a small portion of the current network. However I can't see the changes in government policy and public perceptions need for rationalisation of the network coming to pass any time soon. So the taxpayer gets stuck with yet another white elephant

Another question worth asking, as partial privatisation is on the table, is, Why don't mixed ownership firms do as well as fully privately owned firms? Aidan Vinning and Anthony Boardman's "Ownership and Performance in Competitive Environments: A Comparison of the Performance of Private, Mixed, and State-Owned Enterprises", Journal of Law and Economics vol. XXXII (April 1989) concludes 'The results provide evidence that after controlling for a wide variety of factors, large industrial MEs [mixed enterprises] and SOEs perform substantially worse than similar PCs [private corporations].' The basic problem is that partial government ownership politicises the firm.

As to the performance of state-owned banks Marcio I,. Nakane and Daniela B.Weintraub and look at Bank privatization and productivity : evidence for Brazil. Their abstract reads:

Over the past decade, the Brazilian banking industry has undergone major and deep transformations with several privatizations of state-owned banks, mergers and acquisitions, closing down of troubled banks, entry by foreign banks, and so on. The purpose of this paper is to evaluate the impacts of these changes in banking on total factor productivity. The authors first obtain measures of bank level productivity by employing the techniques due to Levinsohn and Petrin (2003). They then relate such measures to a set of bank characteristics. Their main results indicate that state-owned banks are less productive than their private peers, and that privatization has increased productivity.

Rafael La Porta, Florencio Lopezde-Silanes and Andrei Shleifer also look at the Government Ownership of Banks. They write

In this paper, we investigate a neglected aspect of financial systems of many countries around the world: government ownership of banks. We assemble data which establish four findings. First, government ownership of banks is large and pervasive around the world. Second, such ownership is particularly significant in countries with low levels of per capita income, underdeveloped financial systems, interventionist and inefficient governments, and poor protection of property rights. Third, government ownership of banks is associated with slower subsequent financial development. Finally, government ownership of banks is associated with lower subsequent growth of per capita income, and in particular with lower growth of productivity rather than slower factor accumulation. This evidence is inconsistent with the optimistic development' theories of government ownership of banks common in the 1960s, but supports the more recent political' theories of the effects of government ownership of firms.

This however has to be the coolest paper ever on the trying to workout if there is a difference between the performance of government-organised production versus privately organised production. Jonathan M. Karpoff, "Public versus Private Initiative in Arctic Exploration: The Effects of Incentives and Organizational Structure", Journal of Political Economy, February 2001, v. 109, iss. 1, pp. 38-78.

Karpoff's paper exploits a very interesting and unique natural experiment to compare the performance of government-organised versus privately organised production. Karpoff studies a comprehensive sample of 35 government-funded expeditions and 57 privately funded expeditions to the Arctic from 1818 to 1909 seeking to locate and navigate a Northwest Passage, discover the North Pole, and make other discoveries in arctic regions. I guess these are the cold hard facts!

He finds that the private expeditions performed better using several measure of performance. Karpoff shows that most major arctic discoveries were made by private expeditions, while most tragedies - in terms of lost ships and lives - were on publicly funded expeditions. Karpoff notes that the public expeditions might have had greater losses because they took greater risks, but then the public expeditions would have had a greater share of discoveries, which did not occur.

Karpoff also estimates regressions explaining outcomes in several ways-crew deaths, ships lost, tonnage of ships lost, incidence of scurvy, level of expedition accomplishment - controlling for exploratory objectives sought, country of origin, the leader's previous arctic experience, or the decade in which the expedition occurred. In essentially every regression, the dummy variable for private expedition is significant, with a sign indicating that the private expedition performed better. Karpoff concludes that the incentives were better aligned in the private expeditions, leading to systematic differences in the ways public and private expeditions were organised. While the uniqueness of the sample limits its generality, Karpoff provides an interesting illustration of the impact of ownership on the performance of an organisation.

While it is true that the effect of changes in ownership on the performance of firms is still debated in some quarters, most of the evidence suggests that firm performance improves when SOEs are privatised. This is a result that should be kept in mind with thinking about the current discussion on the merits of privatisation.