Bibek Debroy: Dogs and cows

OF THE RECORD

There are diverse indicators to measure economic development""per capita income (PPP or otherwise), HDI, GDI, GEM.

There are also various indicators to measure, not economic development, but governance. Governance measurement is less robust than measuring economic development, because we lack consensus on what governance means, even if we talk about it all the time.

Accordingly, the same country may have different rankings in different governance surveys. Jeffrey Sachs recently produced a book titled The End of Poverty, reviewed in the Indian media.

The Sachs argument is one of increasing official development assistance (ODA) or aid. While the merits of that argument are debatable, there is a chapter in that book titled "Myths and Magic Bullets".

In that chapter, Sachs tries to counter the argument that aid will not work in Africa because of governance problems. A quote is relevant.

"The point is that virtually all poor countries have governance and corruption indicators that are below those of the high-income countries. Governance and higher incomes go hand in hand not only because good governance raises incomes, but also, and perhaps even more important, because higher income leads to improved governance. As a country's income raises, governance improves for two major reasons. First, a more literate and affluent society is better able to keep the government honest by playing a watchdog role over government processes. Newspapers, television, books, telephones, transport, and now the Internet, all of which are vastly more available in rich countries, enhance this watchdog function and empower civil society. Second, a more affluent society can afford to invest in high-quality governance. When governments are backed by ample tax receipts, the civil service is better educated, extensive computerisation improves information flows, and the public administration is professionally managed." This point is not debatable.

First, is there a strong correlation between per capita income and governance indicators? The correlation is not as strong as proponents of governance make it out to be.

Yes, the correlation is strong for some indicators of governance, but not for all. Second, does this correlation establish causation, and if so, in what direction is the causal relationship?

Does better governance lead to a benign effect on per capita incomes? Do higher incomes lead to governance improvements?

Or is the correlation entirely spurious and are there other factors that lead to improvements in both economic development and governance? We don't have clear answers.

Indeed, apparently bizarre indicators often tell us quite a bit about economic development. We had a livestock census in 2003.

The cattle population is divided into a crossbreed/exotic category and an indigenous category. There are 4.945 million male crossbreed/exotic cattle in the country and 24.686 million of the female variety.

There are 77.534 million male indigenous cattle and 185.181 million of the female variety. It is plausible that the gender (sex) ratio of indigenous cattle is more prone to natural selection, while that of crossbreed/exotic cattle is more prone to human selection.

For crossbreed/exotic cattle, the female figure is 4.99 times the male figure, while for indigenous cattle, the female figure is 2.39 times the male figure.

Clearly, for crossbreed/exotic cattle, we have a problem of missing bulls, if that is the expression to use. This becomes even more interesting if one makes inter-state comparisons.

By any indicator, Bihar is a backward state. Consider the sex ratio for indigenous cattle.

In Bihar, the female/male ratio is 2.56 and it is 2.03 for Orissa, another backward state by any indicator. But in Punjab, the ratio is 7.06. It is 3.76 in Haryana.

Such varying ratios have an obvious link with what the cattle are used for. According to the livestock census, male cattle are used for breeding, work, or breeding and work, and female cattle are used for breeding and milk.

Breeding is a separate issue. But when agriculture gets mechanised, you no longer prefer male cattle for work. Similarly, when agriculture diversifies into dairy, you prefer female cattle.

Ipso facto, unlike the sex ratio for humans, the sex ratio for cattle is an extremely good indicator of the state's level of economic development, more specifically, its level of agricultural development.

Indeed, the cattle sex ratio (CSR) is likely to have a higher correlation with a state's level of economic development than any measure of governance in the state.

There is another indicator, also from the livestock census, which is even more bizarre. This is the population of dogs, although in using this, we have problems with some states in the North-East.

But consider mainstream states. The census divides dogs into domestic dogs and others. We do have enumeration problems, because other dogs are not always strays, or even feral.

Plus, there is a rural versus urban issue. Ignoring those problems, whenever development occurs, barring countries where there are religious bars on keeping dogs or countries where dogs are eaten, dogs move from the streets and into households as pets.

Accordingly, the domestic/other ratio is 4.71 in Goa and 4.11 in Kerala. But it is 1.8 in Bihar and 0.74 in Orissa. This dog transition ratio (DTR) is also a pretty good indicator of a state's level of development.