Monday, December 12, 2011

Economic Habits of Intelligent Countries and Long Term Economic Growth

Introduction

We all know individuals differ in their ability to find the correct answers to objective problems and it turns out that that the average ability across countries differs a lot too. I can therefore compare the differences in policy choices and behaviors across countries to their mean IQs and use the results as the Smart Vote with respect to those policies.

The mean IQs of countries were obtained from Lynn and Vanhanen’s estimates1,2. The national IQ scores had very high test-retest reliability. They are also showed a very high validity in that they correlated highly with country performance on international science and math rankings1,2,3. National IQ differences are associated with a large variety of developmental outcomes e.g. national income, poverty rates, economic equality, economic growth, life expectancy, infant mortality, AIDS rates, homicide rates, democracy, good governance, patent rates, number of scientists, Nobel Prize rates, and many more1,2,3.

Focusing on economics we see that cognitive skills play a large role and that it is measured high level skills and not merely rates of school and university enrolment that counts6. National IQ scores are an extremely robust measure of human capital. Between a quarter and a half of the global income distribution is explained by a single factor – the effect of large, persistent differences in national average IQ on the private marginal product of labor4. . Each IQ point difference is associated with a persistent 0.11% difference in annual GDPpc growth5. That means a relative doubling of income of the top 10% and bottom 10% of countries by national IQ, every 32 years.

Still cognitive capital isn’t everything. The consensus was that the efficiency with which capital is used plays as large a role as capital itself but doubt has been cast on this view7.

Economic freedom is something that could play a role in the efficiency with which capital is used and it does appear to explain some of the variation in national income not accounted for by national IQ2. The thing is levels of economic freedom aren’t just randomly distributed across countries, they’re chosen by their citizens and politicians. Since economic freedom is a choice, and may also be a good idea, the degree of economic freedom in each country should depend on the intelligence of the people.

In this post I will investigate the Smart Vote on various aspects of economic freedom and the role intelligence and economic freedom play in long term economic growth. I use various measures of economic freedom covered in “Economic Freedom of the World” by Gwartney and Lawson.

“Economic Freedom of the World” has an overall Economic Freedom Index, 5 sub-indices - Government Size, Rule of Law, Sound Money, Freedom to Trade and a Regulation Index (made up of 3 further subdivisions – credit regulation, labor regulation and business regulation). A total of 45 individual measures, spanning 30 years, go into it.

I constructed a 30 year average for each of these measures and indices. The point of doing that is because temporary variations are cancelled out leaving a far more reliable and meaningful measure, and also because economic freedom most likely works over long periods. Single year economic growth figures are highly unstable too. For example the average correlation between the growth rates of any two years in the last 30 is only 0.07, meaning that the growth rates of any two years have less than ½ a percent in common. Using long term growth figures is also much more reliable and meaningful. The growth rate over the full 30 years has a reliability of 0.68 – still not great but much higher than 0.07.

I will compare these more persistent differences in economic freedom with the growth rate over the same period as well as with national IQs, while controlling for national income. I control for income because wealth can effect what policies a country is able to choose, and because poorer countries grow faster.

I also look at whether these economic freedom measures do all measure a single variable that we might call “economic freedom”. If the variables seem to tap several different concepts, it doesn’t make sense to talk about economic freedom. It turns out that a single construct does underlay all the measures, but a few measures do not align as the authors would have it. High government consumption expenditure, high transfers and subsidies and the imposition of standards on business turn out to be measures of economic freedom rather than its lack. I extract an economic freedom factor that puts these variables the right way around.

The Smart Vote on Economic Freedom

The first thing to establish is whether average national intelligence has any relationship to economic freedom. Looking at the left hand of the table in Appendix I the answer is “definitely yes.” National IQ is significantly correlated to 80% of the economic freedom measures and furthermore the relationship tends to be strong. Indeed intelligence seems to explain most (59%) of the differences in economic freedom across countries.

Now wealthy countries, which tend to have higher national IQs, could simply be indulging in the ‘luxury’ of economic freedom, and the relationship we are seeing with IQ may be just along for the ride. Controlling for income tests this. I found that income certainly does play a role but by no means does it explain away the intelligence connection.

Being a poor country (whether the population is smart or not) makes providing some forms of economic freedom difficult e.g. ensuring sound money, private and foreign banking, or that black market rates stay close to official rates. A small economy will find it more difficult to manage the value of its currency in the storms of a large global market, no matter how carefully it tries. The economic ‘freedoms’ of low government consumption and not requiring unemployment insurance are also in this category, but this is the default in poor countries simply because they can’t afford them.

On the other hand ensuring that government isn’t a nuisance to business (low regulation of business), not hindering foreign investment or capital movements (including citizens having foreign currency accounts), and making compliance with trade rules cheap, are all genuinely a function of smarter countries being more likely to choose these policies.

Ensuring the rule of law and low corruption levels reflects a mixture of income and IQ effects. Rule of law is costly, so being richer helps, but nevertheless being smarter makes it much more likely.

So the Smart Vote is very strongly in favor of ensuring economic freedom – particularly in the form of the rule of law, the ease of doing business, and the freedom of capital movements and international trade. This is not to say that the other economic freedoms of sound money and freedom of the financial sector are not also intelligent choices – it’s just that success in providing those freedoms is, like it or not, proportional to the country’s economic clout.

The Smart Vote and Opinions on the Merits of Low Economic Freedom i.e. Communism.

Let’s look at the Smart Vote on virtues of economic freedom from the opposite end. Much of the world was once communist, and many intellectuals in the non-communist world were in favor of it. Could it be that we are missing something here? Could it be that the lack of economic freedom in certain forms is also a more intelligent alternative?

The graph below shows the historic trend of the Smart Vote on communism in the USA. In the 70s (and probably before) the Smart Vote was for communism being at least OK for some countries, if not actually good. The Stupid Vote was for communism being the worst possible system, with “Bad but Not the Worst” a somewhat smart choice in between the two.

Many intellectuals are attracted to the idea of a society planned by the likes of themselves, and communism/socialism seemed to be a tradition that presented that opportunity. It also seemed obvious at the time that it wasn’t very clever to allow society to develop in the seemingly random directions a free market would take it.



20 years later the Smart Vote is very different. Viewing communism as the worst possible system is still a stupid view, but now viewing it as good, or even OK, is even dumber. The Smart Vote is now that communism is definitely pretty bad, even if it isn’t the worst possible system.

So why did the Smart Vote ‘change its mind’? The change is entirely due to the more intelligent changing their minds. Those with IQs less than 85 didn’t change their views at all. The reason for the change wasn’t the obvious failure of the bulk of the communist system in the late 80s. It must have been a form of knowledge that gradually made itself apparent from the start because the Smart Vote in favor of communism has declined at a steady constant rate from the 70s and didn’t slow down after the collapse. I suspect it was the steady and strong improvement of non-communist economies. Free markets obviously weren’t drifting anywhere bad. Another factor was probably theoretical developments. Early in the 20th century Marxism was a well developed intellectual tradition with a lot of seemingly smart things to say about economics. In contrast pro free market economic theory started as a relatively poor intellectual cousin, and only developed gradually as a serious competitor. Perhaps it also gradually dawned on intellectuals that communism was a form of social engineering that required a fairly permanent state of brutality to operate.

I also looked at the international trends in economic freedom broken down into quintiles by national IQ. The economic freedom levels of ex-communist countries were added. Ex-communist countries average around the 4th quintile in national IQ. Smarter countries persistently maintain freer economies, but everyone is learning that freedom is better. The change mirrors the change in the Smart Vote on communism in the US. Note the massive change in the economic freedom of the ex-communist countries themselves. They moved from levels lower than those maintained by the most foolish of countries – by this measure they were the most foolish countries, even if they weren’t the least intelligent - to economic freedom typical of their national intelligence levels.



So the answer to the question is a definite no. Communism isn’t, and never was, an intelligent alternative, even if it did take time to realize that. The Smart Vote is for greater economic freedom, and there don’t appear to be any systems that provide a viable alternative for smart opinion.

The Relative Contribution of Intelligence and Economic Freedom to Economic Growth.

The Smart Vote implies that establishing a high level of economic freedom is the ‘correct’ thing to do. In this section I test this idea. Do higher levels of economic freedom per se speed up growth, or is faster economic growth just a function of having a smarter population (high intellectual or cognitive capital). Perhaps economic freedom is purely an indulgence of the smart. Perhaps economic freedom is irrelevant to economic development but it’s just ‘correct’ to have for other reasons – after all Amartya Sen pointed out that freedom is good for its own sake quite apart from any indirect good it may help to establish.

The details of the analysis are in the table in Appendix 2. The effect of each individual measure tends to be modest compared to the role of intelligence but when added together general economic freedom has a bigger impact on growth than intelligence. A large part (65%) of the effect on growth lies in what intelligence and economic freedom have in common. 88% of the role intelligence plays in producing faster growth lies in the role it plays in providing overall economic freedom and only 12% via factors that aren’t part of freedom. On the other hand 28% of the effect of economic freedom on growth has nothing to do with intelligence.

The existence of the Rule of Law is an important factor in allowing economic development, particularly when applied to economic activity e.g. the protection of property rights and freedom from bribery, etc. Providing rule of law is expensive, which is why there is a tendency for greater state consumption expenditure to go with higher growth rates. Earlier I established that it is smart to ensure the rule of law. Faster economic development is a major reason why it smart.

Another pro growth factor falling (with rule of law) under what may be called ‘providing stability’, is sound money, or stable prices and positive interest rates. This is a good reason for assuming that the Smart Vote would be for stable money values when the size of the economy allows it.

The next group of pro growth factors relate more directly to economic activity.

Firstly, growth is faster if government does not make itself a nuisance to business, by burdening it with lots of administration requirements, making it onerous to start a business or costly to comply with tax.

Secondly, economic development is faster when labor regulation is more flexible.

Thirdly, growth is faster when foreign investment is freely allowed and international trade is robust and not hindered by procedural barriers.

Fourthly, growth is faster when resources are not taken away from the more productive - in high top marginal taxes; or given to the less productive - in transfers and subsidies.

Finally, requiring business to meet a certain level of standards seems to speed up growth – probably by reducing waste. (Remember I found requiring higher standards to be pro freedom in my factor analysis. Gwartner & Lawson have not requiring standards as pro freedom.)

The Smart Vote is indifferent to, or even opposed to, some of these factors. For example, higher transfers seem to be a smart thing to do in spite of slowing economic growth. The same applies to not imposing minimum standards. Flexibility on labor regulation seems to have intelligent reasons opposing it that match economic growth in importance. I will refrain from speculating on what these anti-growth reasons may be.

The equation predicting how many times richer a country got over 30 years is as follows

Growth = 0.1106*IQ-0.0005*GDPpc (at start)+1.6913*EF (factor score)-2.3738.

It explains 46.7% of the national differences in the rate of economic growth. It sounds like there is lots of growth potential left to be explained but there isn’t really. Unreliability in the measurements accounts for much of it. For example the equation as a whole has a similar reliability to national IQ and the economic freedom index (0.94 and 0.96) and 30 year growth has a reliability of 0.68. Correcting for that gives a correlation of 0.846, which explains 71.5% of the differences in 30 year growth rates. At most 28.5% of the variation in growth rates (less than a third) has something to do with factors other than national intelligence, current income and economic freedom. Some of this is just plain luck.

The factor score is in unit normal scores. The difference between being in the top and bottom 10% of economic freedom (Australia and Norway versus Cameroon and Ecuador say) is equivalent to a national IQ difference of 23 points. In individuals an IQ gap of that magnitude makes the difference between being virtually certain of being able to solve a tough problem, and virtually certain of not being able to solve it. Economic freedom is therefore a way of massively increasing the efficiency of a country’s cognitive capital. Hayek often made the point that free markets use knowledge incomparably more efficiently than command economies. The equation both confirms that view and gives it a slightly different slant.

One gets another perspective on the impact of such a difference in economic freedom by looking at the likely differences in economic development. The median country more than tripled its per capita income over 30 years - 3.4 fold to be exact. For the bottom 10% of economic freedom it would likely be 2.12 and for the top 10% it would be 4.68 fold. In other words if two countries started with the same income and national IQ, where one chose to maintain a bottom 10% level of economic freedom, and the other a top 10% level, the freer country would have more than twice the income of the less free country within 30 years.

Conclusion

The results are a vindication of the Smart Vote concept. The policies indicated by the Smart Vote as most likely ‘correct’, proved to be a very important component of economic development. These policies were to provide economic freedom through the provision of stability via the rule of law and sound money, and through leaving business alone to hire, produce and trade as best they see fit,

The Smart Vote isn’t always right. It was wrong about communism in the 70s. On the other hand it is very sensitive to changes in the state of knowledge, and very likely makes the best possible decision given the state of knowledge at the time.

References

1. IQ and the Wealth of Nations, Lynn R. & Vanhanen T.
2. IQ and Global Inequality, Lynn R. & Vanhanen T.
3. The g Factor of International Cognitive Ability Comparisons: The Homogeneity of Results in PISA, TIMSS, PIRLS and IQ tests Across Nations, Rindermann H.
4. IQ in the Ramsey Model: A Naïve Calibration, Jones G.
5. Intelligence, Human Capital and Economic Growth: A Baysian Averaging of Classical Estimates Approach, Jones G & Schneider J.
6. The Role of Cognitive Skills in Economic Development, Hanushek & Woessmann.
7. Accounting for Cross-Country Income Differences, Francesco Casseli.

APPENDIX 1

Out of the 45 economic freedom measures, national IQ is significantly correlated to 36 of them (80%) and is close on 2 more. Of the 9 variables where it isn’t, 6 are from one area – labor regulation. Furthermore the relationship tends to be strong. Indeed if the general economic freedom factor I extracted is anything to go by, intelligence seems to explain most (59%) of the differences in economic freedom across countries.

GDP per capita is significantly related to 26 of the 45 economic freedom measures. Mostly, what intelligence and income have in common is what matters for economic freedom. That means is that either the apparent effect of intelligence is a by-product of its close relationship with income, or the apparent effect of income is a by-product of its close connection with intelligence, or some mixture of both. To tell which is which, compare the “IQ alone” and “GDP alone” columns. If, for example, the “IQ alone” number is high and the “GDP alone” is “ns” (not significant), then any apparent income effect is because income rides on the shirt tales of intelligence. If the numbers are the other way around then economic freedom depends on high income and the IQ effect is not real.



APPENDIX 2

29 of the 45 economic freedom variables are associated with economic growth beyond the connection to intelligence. The effect of each tends to be modest compared to the role of intelligence but economic freedom in general has almost three times the independent impact on growth as intelligence. In fact a considerable part (88%) of the role intelligence plays in producing faster growth lies in the role it plays in providing economic freedom.

1 comment: