« John Berry: Greenspan Sticks to His Guns, as Well He Should | Main | My First Day of Class »

Tuesday, September 25, 2007

"Which Institutions Matter for Economic Growth?"

How do institutions interact with economic growth? This research finds that, at least in the case examined, South Africa, the establishment of property rights is an important factor in unleashing economic growth. However, given the warnings of Dani Rodrik, I'm hesitant to generalize too much from this result. Dani's argument is that every country is unique in its institutional and social design so that each will also have it's own unique choke point that is inhibiting growth. For some countries it may very well be private property rights, but for others it may be something else entirely that needs to be changed to allow growth to accelerate. Nevertheless, whether it's identified a broader principle or not, this research is helpful because it tells us that when a country has conditions similar to those that existed in South Africa, the establishment of private property rights is a good place to start when considering what institutional reforms to implement:

Which institutions matter for economic growth?, by Liam Brunt, Vox EU: It is obvious that a country’s political, legal, economic and social institutions will affect its rate of economic growth. However, it is much more difficult to identify exactly which institutions matter and exactly how they matter. This is an issue of some practical importance. Countries are free to redesign their institutions in order to improve their economic performance. But, unless they can pinpoint the beneficial aspects of particular institutions, the only option is to import wholesale the institutional structures of another, more economically successful country. This happened in Japan in 1945 with respect to many US institutions and again recently in Dubai, which adopted the entire panoply of commercial law that regulates the City of London. However, in many cases it may be infeasible or inefficient to change the entire institutional régime, or it may be politically or socially unacceptable. For this reason, it would be useful if we had a better idea of exactly which aspects of which particular institutions were beneficial for stimulating growth. But the evidence on this matter is very mixed.

When we talk about “institutions”,  we are referring to something much broader than simply the set of easily recognisable legal entities, such as parliaments or central banks or unions (although these are all particular examples of institutions). An institution is any generally accepted procedure that governs the process of interaction between members of a society. For example, waiting in line is a well-known institution for allocating goods that is particularly popular in England. Property rights are a fundamental legal and economic institution, although the concept of a property right and the ways in which it might be regulated or enforced vary greatly across the world. Legal systems are another fundamental institution and here much attention has focused on the distinction between civil law and common law systems – the former being based on a set of statutes handed down by a supreme authority (such as the Code Napoléon), the latter being based on a corpus of case law formulated by judges (such as the English common law system). A second characteristic of legal systems that has attracted attention is the degree of independence of the judiciary from the legislative authorities (such as whether Supreme Court judges have to be reappointed regularly by the legislature). An important social and economic institution that is distinct from the legal system, yet closely related to it, is that of corruption; corruption provides an important and accepted framework for economic transactions in many countries, although is virtually absent from some others.

A challenge that we face in disentangling the economic impact of these and other institutions is that their occurrence tends to be very highly correlated across countries, due largely to the impact of colonial inheritance. For example, the English legal system is based on common law and the English governmental system is based on parliamentary democracy; therefore the former English colonies virtually all have a common law system and a form of parliamentary democracy. By contrast, the French and Spanish legal systems are based on the Code Napoléon and – at the time of colonisation – the French and Spanish governmental systems were quite autocratic; hence most former French and Spanish colonies have a civil law system and more autocratic governmental institutions (such as a prominent role for the Presidency). To make these various threads even more difficult to disentangle, the institutional factors may be correlated also with geography. For example, England colonised many territories in the temperate zones (North America, Australasia and southern Africa) whereas France and Spain colonised many territories in the tropics (West Africa, together with Central and much of South America). More worryingly, it has been argued that geography itself systematically influenced the type of institutions that were bestowed upon a colony, instituting more extractive forms of government in regions where there was a large indigenous population to be exploited. Yet geography can have a large impact on current economic performance in its own right, due to positive factors such as natural resource endowments (mineral deposits and so on) and negative factors such as the disease environment (notably the prevalence of malaria). Thus the correlation between geography and institutions makes it difficult to estimate the separate effects of each of them on economic growth.

Instead of looking across countries, where these correlations are problematic, an alternative approach is to consider the performance of one country over time. Since geography is held fixed for any particular country, it cannot contaminate any analysis of institutions. But we need to find a country that has significantly changed its institutions over time so that we can consider its performance before and after the changes; and we need to have sufficiently detailed data that we can reliably detect any performance changes. We also need to be sure that the performance changes occurred in response to the institutional changes, rather than vice versa. South Africa meets all these requirements. It was founded by the Dutch East India Company in 1656 as a stopover on the trade route to the Far East. The colony was seized by the British in 1795 for geo-political reasons (i.e. to stop it falling into the hands of the French during the Revolutionary Wars); it was given back in 1801 by the Peace of Amiens; and then seized again (definitively, it later turned out) in 1806 during the Napoleonic Wars. The Dutch system of property rights, law and administration were followed continuously until 1814; although the British had become the new overlords in 1795, they had no long term plan for the colony and therefore no interest in reforming the colonial institutions. But when British tenure became permanent in 1814, they undertook active steps to strengthen the colonial economy (such as improving the food marketing system, awarding prizes for agricultural innovation and reforming the labour laws that governed the employment of the black population). In 1827, the British imposed the English common law system over the former Dutch system and sent the judges to be re-trained in England. In 1843, the British introduced a new Empire-wide policy that required public lands to be auctioned off in freehold to settlers; previously, most land in South Africa had been held from the government on fixed term tenancies, or else simply by squatting.

Thanks to the extraordinary recordkeeping of the Dutch and British colonial governments, which was far more intrusive and complete than anything that was permitted in Europe at that time, we can trace the effects of each institutional reform on agricultural output. Since diamonds and gold were discovered only in 1867 and 1886 respectively – long after the period that we shall be considering – agriculture constituted the vast majority of South African output up to the 1860s and offers a good guide to overall economic performance. The important breaks in South African output growth are immediately apparent in figure 1.

Figure 1. Annual agricultural output in South Africa, 1701-1875.

There was a substantial step up in output after the British took over in 1795, despite the absence of any formal institutional changes. This boom was perhaps due to the inability of the new British government to interfere in local affairs as effectively as their Dutch predecessors, who spoke the language of the locals and knew most of them personally: an ineffective tax collector translates to a reduction in the effective tax rate, which may have encouraged an increase in output. But South African output was then static for the next 30 years, through two periods of important institutional reform: the government attempt at economic stimulation after 1814; and the transformation of the legal system from a civil to a common law régime in 1827. Only when private property rights in land became substantially more secure in 1843 do we see any significant improvement in the rate of output growth, at which point it suddenly quadruples. There is considerable evidence from agricultural surveys and census data that the increase in output was due largely to very substantial increases in fixed capital investment, particularly the creation of reservoirs and irrigation systems in areas that were naturally too arid for wheat production.

So which institutions matter for economic growth? South African institutional development suggests that government efforts to stimulate growth by revising the institutional framework (such as labour laws and marketing regulations) are largely ineffective in the long run. Nor is it important whether legal disputes are resolved in the context of a civil or common law régime. By contrast, firmly establishing private property rights has very positive effects. This result is perhaps not surprising. Changing particular institutions is really only tinkering at the economic margins. But establishing clear property rights facilitates an almost infinite variety of economic interactions – most of them unforeseen and unforeseeable ex ante – that is limited only by the imagination of economic agents to dream up new types of contracts that permit them to more efficiently exploit the available resources. It is the firm establishment of private property rights that unleashes the full potential of the economy. Several developing economies – such as Vietnam and China – have recently been moving down this road and history suggests that the economic gains are likely to be large.

* For further details, see “Property Rights and Economic Growth: Evidence from a Natural Experiment,” Liam Brunt, CEPR DP 6404. September 2007.

    Posted by on Tuesday, September 25, 2007 at 12:33 AM in Economics | Permalink  TrackBack (0)  Comments (48)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference "Which Institutions Matter for Economic Growth?":


    Feed You can follow this conversation by subscribing to the comment feed for this post.