Avinash K Dixit: Governance and development

Corruption is only a part of the more general issue of economic governance, and of the design and reform of institutions of governance.

The problem of corruption has become very important in recent thinking on development policy. For example, the World Bank's governance web site, http://www.worldbank.org/wbi/governance, seems to focus almost exclusively on corruption. In the Indian context, observers like Luce (2007) identify it as a key issue affecting the country's growth prospects. But corruption is only a part of the more general issue of economic governance, and of more general policy questions concerning the design and reform of institutions of governance. Economic governance comprises many organisations and actions essential for the good functioning of markets, most notably, the protection of property rights, enforcement of contracts and provision of physical and informational infrastructure.

The importance of secure property rights can hardly be overstated. Without them, people will not create or improve the assets, physical and intellectual, that are essential for economic progress. De Soto (2000) builds the argument and marshals the evidence in a thorough and compelling book. Field (2006) has taken the case even further. The security of property rights not only increases the supply of capital and efficiency in its allocation; it also increases labor supply. When titles to land and capital are official and secure, people need not spend time and effort to guard their rights, so they can put the labour and time to productive uses. Field's empirical research on the titling programme in Peru finds large and significant effects: "For the average squatter household, property titles are associated with a 14 per cent increase in household work hours, a 28 per cent decrease in the probability of working inside the home, and a 7.5 per cent reduction in the probability of child labour among single-parent households. Panel estimates ... support the cross-section results: between 1997 and 2000 household labour supply increased an additional 13 hours per week for squatters in neighborhoods reached by the programme."

The biggest threat to property rights in many countries comes not from private individuals who exploit the weakness of government institutions, but from the government itself and its agents. These threats need not take the form of outright theft. In fact, confiscation or nationalisation without compensation has been relatively rare for the last four decades. I want to emphasise that the big problem is the unexpectedness and arbitrariness of these actions of the government and its agents. A predictable tax or a predictable level of corruption will also deter economic activity, but the deterrent effect of uncertainty is likely to be much bigger. How can people ensure that their government and its agents respect property rights? Waiting for the government to eliminate corruption may be futile. Individuals can do little; collective action is needed. Greif, North and Weingast (1994) show how merchant guilds in medieval Europe performed this function to keep monarchs from expropriating foreigners trading in their realms. Perhaps confederations of industries in modern economies can perform similar functions.

The second finding I want to highlight is the distinction between organised versus disorganised, or unified versus non-cooperative, corruption. If a project needs nineteen permits issued by nineteen separate licensing and regulatory agencies, each of them can try to extract as much as they can from the applicant, not taking into account the fact that the implied tax levied by each of them discourages the activity and thereby reduces the take of all the others. If all nineteen permits are in the hands of one agency, it will recognise this interdependence and therefore will impose a lower tax, that is, engage in less corruption. This argues for the creation of "one-stop" licensing and regulation authorities for each kind of economic activity.

Finally, consider a country that is introducing a modern and formal system of titling as De Soto and others would recommend. They are not doing this with a completely clean slate. Most societies without formal legal titling have some traditional system of rights, determined by tribal chiefs or village elders or heads of extended families, and enforced by these traditional authorities using various systems of social norms and sanctions. These rights may not work perfectly, but they exist, and will interact with the formal rights that are being introduced. If this interaction is dysfunctional, the formal rights may not work as western advisers would wish. Ensminger (1997) found just such a problem with land rights in Kenya. The traditional system guaranteed shares (usufruct rights) to various members of the extended family of the purported owner. This made it infeasible to use the land as collateral in a loan application from a formal sector bank, thereby defeating one of the most important advantages of titling offered by De Soto.

Kranton and Swami (1999) found that the introduction of civil courts in colonial India interacted adversely with agricultural credit markets in just this way. Competition among lenders increased. But traditionally lenders used to reduce risk for farmers by subsidising their investments in times of crises; they could no longer do so because the courts enforced only simple debt contracts, not complex contingent risk-sharing ones. The overall outcome was a worsening of social welfare. These examples bring out the importance of ensuring that new formal systems relate synergistically, not adversely, with the informal and traditional systems.

I think that waiting for the political process to institute the needed reforms "top down" would be a mistake. Even something as basic as the periodic publication of companies' audited accounts was initiated privately by J P Morgan when he started Federal Steel with Elbert Gary in 1898, because they believed that "corporations issuing publicly traded securities had to account for their financial performance" (Strouse 2000, p 398). Later this principle was taken over and implemented in legislation by the progressive movement under Theodore Roosevelt. And recently the CEO of Aflac, an insurance company, has voluntarily allowed the shareholders a vote on his compensation; this may also spread. Institutional investors can similarly play a major role in improving corporate governance; in the US it is said that the California public employees' pension fund CALPERS has been more important than the Sarbanes-Oxley Act in this way.

In dealing with corruption, shining light on corrupt activities and exposing them to fresh air may be the most important starting point. In this respect India has the great benefit of a free press; no government and no media mogul should be allowed to depreciate this asset. The Right to Information Act can have major beneficial effects by removing information asymmetries and improving accountability. On the whole I believe that bottom-up and organically generated reforms will work better than imposed top-down ones.

The author is John J F Sherrerd University Professor of Economics at Princeton University. This is an edited version of the PR Brahmanda Memorial Lecture he delivered at the RBI on June 28, 2007. The full version can be accessed at http://www.rbi.org.in/scripts/BS_SpeechesView.aspx?Id=343