Latitude and Income per Capita in Comparative Development

NOTE: The Growth Economics Blog has moved sites. Click here to find this post at the new site.

New paper out by Holger Strulik and Carl-Johan Dalgaard (who I predict is at this moment taking a smoke break). The paper looks at the reversal of the latitude/income relationship over history, and propose a physiological reason for it.

For starters, if you are familiar at all with basic development statistics, then you probably know that latitude and income per capita are positively correlated. The farther away from the equator you get (higher latitude) the richer you get. This works going north or south. South Africa is richer than Nigeria, for example, and Chile is richer than Ecuador. Dalgaard and Strulik have a nice graph showing this relationship holds not only for all countries, but also within Europe.

Dalgaard and Strulik 2004

The first really interesting fact in the paper is that this gradient reverses if you look at pre-Industrial Revolution data. For 1500 CE, there is a negative relationship, and countries that are closer to the equator are richer. Again, this also holds within Europe. I had some vague concept that it probably reversed in the whole sample, but the within Europe evidence is really fascinating.

Dalgaard and Strulik 2014

Around 1500, the Mediterranean countries in Europe were better off compared to their Northern neighbors. An aside: Were the Greeks and Italians of 1500 tsk-tsk-ing their profligate Teutonic cousins for their lazy attitudes and lack of robust economic institutions? Discuss.

Anyway, the latitude/income reversal, and the fact that it holds up within Europe, are both by themselves the kinds of stylized facts that you should cram into your head when thinking about comparative development.

But given that you have crammed that information in there, you probably have several questions. (1) Why are hot places rich in 1500, and cold places poor? (2) What changed to make the cold places rich today?

Dalgaard and Strulik take a swipe at these questions, focusing on the physiology involved in hot and cold places. There thesis rests on “Bergmann’s Rule”, which is a biological regularity noted in 1847. Bergmann’s rule states that average body mass of organisms rises as they get farther from the equator. This holds for people as well as animals. People generally have higher body mass farther from the equator (and no, that’s not just because of Wisconsin. I kid. Sort of.).

Why does Bergmann’s rule hold? Surface area to mass ratios. Big people have lower surface area to mass ratios, so they are more thermally efficient in cold climates. Thus the optimal body type for high, cold, latitudes is large, while for places closer to the equator small body types are optimal to maximize surface area to mass in order to radiate heat.

Now, large bodies have an additional feature. They require a lot of fuel (food), in particular for mothers when pregnant. Big women having big babies means using a lot of food. Thus people in cold latitudes were able to have fewer babies, given a supply of food, than their peers in equatorial regions. So we have bigger populations in equatorial regions and smaller ones in cold latitudes prior to the IR. Big populations mean more innovation in almost any type of growth model you write down, so equatorial regions had more innovation during the pre-IR era, and hence were richer.

But, eventually even the cold latitudes are going to innovate far enough to get the point of inventing technologies that rely on human capital. And the cold climate physiology gives them a natural tendency to favor quality over quantity of kids. Thus families in higher latitudes are going to more easily adopt the human capital using technology. This then starts a feedback effect, where by having a few, high-education kids means they can use the human-capital technology. Which raises income per capita. Which leads to further investment in kids at the expense of family size, and cold latitudes enter the Demographic Transition ahead of equatorial regions.

The reversal is inevitable in their model, given the initial physiological difference between latitudes. The physiological story is also consistent with differences in marriage patterns and child birth patterns between Europe and much of Asia in the pre-IR era.

They use Europe as an example, and how the latitude/income relationship holds today. But it holds in the U.S. as well. Is the income per capita of states in the U.S. consistent with the implied physiological differences between different areas of Europe, Africa, or Latin America due to population composition?

This paper predicts a reversal, but this reversal has to happen “just so” to avoid becoming a-historical. That is, the reversal has to happen just before the equatorial countries (China, India, various iterations of Islamic empires) become sufficiently rich to colonize Europe, snuffing out their development. This leaves Europe to effect the reversal, and go out to colonize the rest of the world. Did Europe get lucky here, or is there some reason that those places don’t become colonizers? Luck might be the answer, as you’ve got plenty of close-run things in European history [the Mongols turning back, Lepanto, Vienna].

The last thing that comes to mind here is that for this physiological difference to have such persistent effects, the family patterns it determines must become either (a) genetically rooted into populations or (b) some deeply ingrained in culture as to be permanent. Fertility behavior is mutable. For it to continue to be a reason for lack of development in equatorial regions you need some strong force keeping people locked into the “bad” preference for lots of kids. What is that force?

10 thoughts on “Latitude and Income per Capita in Comparative Development

  1. NYC is on the same longitude as Northern Spain. What do they say about longitude.
    Problem with this latitude thing is it doesn’t take much account of history. Romans came to Britain found them backward; dragged them up by the sandals laces. Romans left Britain and Britain declined for hundreds of years thereafter. It is all about management not latitude.
    And this nuclear family thing is just the end. Plenty of large families in the developed world have large incomes and the children of such families can move on to larger incomes still.

    • That’s why they look at within-continent effects (like within-Europe). Within continents, latitude has some meaning, but across continents, not so much. The real story here isn’t about latitude, per se, but the implied climate. Lower latitudes tend to have more mild climates, but that also depends on other things, like ocean currents. So yes, Spain is same latitude as NYC, but Spain actually has a milder climate. So think of their paper as a “warm-climate reversal” kind of thing.

      • I see from the references that they quote from Professors Acemoglu and Robinson. They are of the institutions are the problem type school. They don’t buy the geography thing which I agree is not strictly the same as climate in this paper. But as climate is the main point it leaves me with why in some countries with such advantageous climates that three crops a year can be grown are they receiving food aid. It is the institutions, I suggest, not the latitude. Previous posts have mentioned the school of get an agricultural base and then develop from there. I think combining the various ideas, true or false, of development related to latitude, geography, agriculture and institutions would make an appropriate post for a development economist or growth economist who might one day be famous. Rather, it is probably a book.

      • Right, so there definitely are institutional issues that prevent equatorial nations from developing. But you have to be careful about making assumptions about agricultural productivity. It’s not true that having a climate for 3 crops a year means they are 3x more productive than a northern climate with only 1 crop per year. For the existing crops/technologies we have, high-latitude temperate zones are vastly more productive than any equatorial zone. This is true even holding technology (tractors, fertilizer use) constant.

        Growing crops in temperate latitudes is low-cost because of regular rain and mild temperatures, as well as yearly frosts that kill all the nasty parasites. In equatorial regions, the parasites never die, rain is sporadic (think monsoons) meaning you have to build massive storage/irrigation systems to grow anything, and temperatures that get too high for too long for lots of crops to grow well.

  2. An interesting version. Actually, we need a theory explaining why something happened after 1500, after all written about post-1800 development.

    Joel Mokyr had a relevant talk a couple of year ago on how an innovation-driven society was born about 1500 ( He naturally mentions population density, but only along other factors, of which communications and opportunity for dissent are important.

    • Anton – my feeling is that you’re right. The “line” of divergence seems to get pushed back further and further to more like 1600 or 1500. There is Mokyr’s work, as well as the newer econ history work on wage series that shows take-off in England and Netherlands earlier than 1700 (at least in leading cities). Makes the IR seem less like the big breakthrough, and more of the moment at which we actually noticed the radical changes taking place.

  3. Pingback: The Reversal of the Latitude-Income Correlation

  4. Pingback: The Reversal of the Latitude-Income Correlation | Homines Economici

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s