Last year, Laveesh Bhandari and I started an exercise of ranking Indian States on economic freedom. Such studies exist at the cross-country level. They are rarer at sub-regional levels and have not been attempted in India earlier.

This is about economic freedom. It is not about political freedom or social freedom. Had we attempted to capture these other categories, as opposed to economic freedom, we would have used different variables.

In the work that Laveesh and I do on Indian states, we have never used responses to questionnaires. Our rankings, including this one, are based on officially published government data. And because such data aren't always available for every state, we couldn't rank all 35 states and union territories last year.

We ranked 20 last year. And this year too, we are stuck with 20. Our 2004 rankings resulted in some controversy in the first half of 2005, chiefly because people hadn't bothered to read what we had actually done. This is a brief preview of this year's rankings.

The entire study will be in the public domain in a few weeks. The time-line for data used this year is mostly 2002-03. In some instances, it is 2004 or 2005. And in rare instances (judicial data), it is 1998-2002. Had later data been available, we would have used that.

Broadly, there are three heads for measuring economic freedom -- size of government (revenue expenditure, administrative component in state domestic product, power subsidy, government share in organised employment, taxes, stamp duties, divestment); legal structure and security of property rights (stolen property recovered, violent crimes, economic crimes, vacant judiciary posts, completion of investigations by police, completion of trials by courts); and regulation of credit, labour and business (minimum wages, man-days lost in strikes and lockouts, share of unorganised labour force, number of SEZs, license fees for traders, market fees, implementation of Industrial Entrepreneurs' Memorandum, power shortages, pendency of cases and persons arrested under corruption-related laws). All variables are naturally normalised.

Rarely have we been criticised for commission of variables. We have often been criticised for omission or non-inclusion of variables.

But in all such instances, when additional variables are suggested, we find that objective data across all states don't exist. Although data are objective, there is minor subjectivity in our rankings. This concerns weights attached to variables, required for aggregation.

However, when we use alternate methods of weighting, we find the state scores don't change that much. The ranks do. This is a point that needs emphasis. People are obsessed with state ranks. What is important is scores that drive the ranks. A difference in ranks may be due to a marginal difference in scores. But it may also be due to a significant difference in scores.

There are, thus, three sets of scores and ranks, one for each head. On size of government, Jharkhand and Gujarat are at the top and Uttaranchal is at the bottom. On legal structure and security of property rights, Tamil Nadu is at the top and Bihar is at the bottom.

Finally, on regulation of credit, labour and business, Kerala is at the top and Chhattisgarh is at the bottom. In reporting scores under these heads, we have reported aggregation of variables using equal weights. Space limitations prevent detailed reproduction of tables under these three heads.

Instead, we report the overall or composite table, obtained by aggregation of scores across the three heads. Tamil Nadu tops the list and Assam is at the bottom. The scores show that Tamil Nadu is at the top by a significant magnitude and Assam (at the bottom) is also a fair distance away from Uttaranchal. Explanation as to why a state performs the way it does, lies in the movement of individual variables.

However, most movements concern legal structure and security of property rights. The impact of size of government is less palpable. And there is little change due to regulation of credit, labour and business, except for declines in Rajasthan and Chhattisgarh.

For comparisons, the table also reproduces the 2004 ranks and scores. One notices significant improvements in Tamil Nadu, Madhya Pradesh, Jharkhand, Uttar Pradesh, Karnataka and West Bengal and significant deterioration in Maharashtra, Rajasthan, Jammu and Kashmir, Orissa and Chhattisgarh.

Indeed, one should flag not only the deterioration in ranks, but the absolute decline in scores in states like Maharashtra, Rajasthan, Jammu and Kashmir, Orissa, Punjab, Chhattisgarh, Bihar, Uttaranchal and Assam.

Governance, and its lack, is an elusive concept and any equation of good governance with economic freedom can be debated. However, economic freedom is about personal choice, voluntary exchange, freedom to compete, protection of person and property and availability of legal and monetary arrangements that not only protect property rights of owners, but also ensure enforcement of contracts.

From this perspective, greater economic freedom should be the objective of governance and on this account, there is no evidence of convergence across India's states.