And there was light… and poverty

Perhaps you had once wondered why all countries within the tropics are poor or almost poor. If you take a look at the map again, you will observe that all of the world’s wealthiest countries, with the exception of the tiny island of Singapore and the sub-nation of Hong Kong, lie within the temperate zone. Of course not all non-tropical countries are rich; but it pretty much appears that year-long sunshine does not go hand in hand with riches. (Too bad for some of us, who would have liked to have it both, sunshine and money.)

The notion that climate affects economic performance, or “environmental determinism” as some people call it, is an age-old concept. The 18^th century French enlightenment thinker Montesquieu was among those who saw a link between climate and human behavior[1]. However, the role of climate on economic performance is out of fashion in current economic thinking, perhaps because of its deterministic and simplistic appearance. Nonetheless, it remains true today as much as ever that poverty is more prevalent in the tropics.

Figure 1 provides a scatter plot and a linear regression fit of average annual temperature on GDP per capita for a sample of 94 countries. The significance of the correlation is remarkable as is its strength: for every rise of annual temperature by 1 degree Celsius, GDP per capita falls by 6%. If Uganda, a country on the Equator, were as cold as Switzerland (a large 17 degrees fall in average temperature), its GDP would double, raising its income to the level of Pakistan. An alternative estimation gives the equally remarkable result that GDP increases by 13% for every one more month with below zero average temperature. The model does over- and under- shoot in predicting the income of some countries. Singapore is too rich for its climate, and Kyrgyzstan should have been much richer given its climate. But it does a good job for many countries; for example the incomes of countries like Canada and Argentina can be almost perfectly predicted once we know their climate.

Is this relationship totally a fluke or is there some sensible reason why income should be lower in the tropics? Thanks to the pure exogenous nature of climate, the standard critique of endogeneity cannot apply here. Could it be that this is purely coincidental? The fact that the economic significance of climate is raised by early thinkers and confirmed by recent data warrants that it deserves a closer look.

I think there are at least two reasons why climate can affect economic performance. Firstly, the potentially heavy winters in the temperate parts of the world could provide unique incentives for innovation and encourage forward looking behavior. If necessity is the mother of invention, I would say the European and northern Asian winter must have begotten a host of inventions[2].

Major examples are of course those innovations that directly tackle the physical needs related to the winter[3]. The radiator, central heating, and many more household items had to be invented for the purpose of surviving the winter. The cold climate of England might have increased the popularity of coal, especially for domestic heating, since the 14^th century. Coal later played the central role of powering steam engines during the industrial revolution. It can be conjectured that the first horse drawn carriages appeared with the aim of avoiding the shrill chill of horse riding in the winter. Several centuries later the carriage was upgraded to the car and the train while the tropical traveler never imagined the need to change the ride of his horse or camel.

The lack of winter in tropical areas meant that the inhabitants were not pressed to develop similar innovations. What is important here is not the innovation itself, but the process of innovation and the culture of systematically solving the constraints imposed by nature. This is not to say that the there are no natural constraints in tropical areas, but I believe that the winter specially stands out as a unique natural challenge with an existential threat. Because of the winter, people have to plan ahead, save a certain fraction of income throughout the year, and invest on items that are required during the winter. It follows that people in temperate climate have adopted a long term oriented perspective to life that emphasizes thrift and a systematic application of knowledge for problem solving.

Secondly, the winter can have a role in the way society is structured. I believe the presence of cold winters provides people with the incentive to cooperate in order to counteract its effects. When I was a young boy growing up in Ethiopia, I sometimes used to run away for a day when I quarreled with my parents. I then went fishing with my friends to the farthest river, or set out to climb one of the nearby mountains. Of course my parents would be even more upset when I returned very late, but I knew that they were also worried. To avoid escalating the case further, they would pretend nothing happened and it was soon forgotten. But a 10 year old Dutch boy in my place could never play the same trick in a winter month. He has to sit down and talk to his parents and resolve the differences. By forcing people to stay inside the house, the winter can bring the family together, and compel them to undertake the challenging task of talking to each other and arriving at compromise.

What evidence is there for the direct effect of climate on culture? Figure 2 provides a scatter plot and a linear regression fit between an index of “long-term orientation” and average temperature. “Long-term orientation” is a cultural indicator that measures the extent to which a society plans ahead in the future. This indicator was developed by Geert Hoefstede and has been used a lot in scientific research.

The figure depcits an unmistakable negative correlation between the culture of long-term orientation and average temperature. The index of long-term orientation, which varies between 0 in Guinea-Bissau (PUE) and 100 in South Korea (KOR), falls by 1.5 percent when average temperature increases by 1 degree Celsius.

Now we can connect the dots and, using the relationships exhibited in Figure 1 and Figure 2, show that average temperature lowers GDP per capita through its effect on culture. Figure 3 shows the relationship between the index of long-term orientation and GDP per capita. The values for the index of long-term orientation used in this Figure are predicted values using temperature. Therefore, this relationship is akin to the standard instrumental variable estimation.

Well, the results are all there to see. Predicted values of the index for long-term orientation have a clearly strong positive effect on GDP per capita. Since we are using only the part of variation of “long-term orientation” that is explained by the average temprature, this is a clear evidence for negative effect of climate on GDP per capita. There is enough here to say: It is the climate stupid!

You may say that this is an awful news for tropical countries. But does the result show that a country’s level of income is preordained by its distance from the Equator? I do not think so. To say that climate matters is not equivalent to saying that economic performance is determined forever as implied by “enviromental determinism”. Let me explain.

Figure 1 shows that more than three quarters of the total variation in GDP per capita remains unexplained. So yes, climate is an important determinants of income, but there is plenty room for other factors too. Picking up the earlier comparison, Uganda could double its income if it were as cold as Switzerland. But the income gap between the two countreis varies by a huge factor of 40; thus even after its income doubled Uganda would remain 20 times as poor as Switzerland. This clearly illustrates that even if Uganda’s climate were perfect, there is much to be done before its income gap is totally iliminated. The point is that there is enough room for imporving economic performance by fixing changeable policy and institutional vriables.

Related to this, Figure 1 illustrates that climate is not an important determinant of income at least in a few countries. Singapore and Hong Kong are worth mentioning again as examples of countries which grew inspite of their tropical climates. The Figure also shows that many countries such as Luxumberg, Malta, and even the USA have larger incomes than their climate would warrant. Many of the new Asian Tigers that are growing ferociously, including Malaysia, Thailand and Indonesia, are tropical countries. This is because, although climate is predestined, clulture can clearly change. People are adaptive and, if given the opportunity, it does not take long for a society to catch on the benefits of long term oriented customs such as thrift and education.

---

[1]  Although, I must admit, the generalizations he makes in his book “The Spirit of Laws” are in many ways overblown, and potentially biased.  

[2] In Aldous Huxley’s book “A Brave New World” the World Controller asks Mr. Watson where he would like to be deported to. “By the way, Mr. Watson, would you like a tropical climate? The Marquesas, for example; or Samoa? Or something rather more bracing?" Mr. Watson replies "I should like a thoroughly bad climate. I believe one would write better if the climate were bad. If there were a lot of wind and storms, for example …"

[3] It is reasonable to conjecture that people in temperate climates were the first to start wearing clothing and shoes, although it is difficult to definitely prove it due to the long time distance ever since. In the Adventures of Tom Sawyer, Mark Twain describes Tom as “the first boy that went barefoot in the spring and the last to resume leather in the fall,” reminding us that even in the 19^th century, shoes were mostly parts of a winter wardrobe.