In a just published paper (in the Proceedings of the National Academy of Sciences [of the United States of America]) on the effects of climate change on growth, Noah Diffenbaugh and Marshall Burke, argue, using a complex model, that the temperature change driven by CO2 emissions has affected poor counties the most, and reduced their growth by a cumulative 17 to 31 percent (between 1961 and 2010). Population-weighted between-country inequality has therefore increased because of climate change. The key result of their study is in a graph (panel B) reproduced below…
which shows that, compared to the situation without climate change, the poorest 10% of the population in the world (more exactly, the poorest decile of world population if people are ranked by their countries’ GDPs per capita) have lost a quarter of their output while the rich countries have gained approximately 25%.
Now, the full model is impossible to figure out from a short text of five pages, but, so far as I can tell, it is based on three key links. First, that increased carbon emissions caused an increase in worldwide temperature. Second, that the increase in temperature is uneven across the world. Third, that the increase in temperature is particularly bad for the countries in the tropics which already suffer from hot climate, and from extreme climate events like droughts, storms etc. As the authors write “historical warming has reduced economic growth and lowered per capita GDP” of poor countries because “the mean temperature of…many poor countries lies in the extreme warm tail” which is too high for economic activity.
Of the three links the most difficult to prove is, I think, the third: that the climate change (more exactly, higher temperature) is responsible for the slowdown in poor countries’ growth (mostly in Africa). Notice that if true, the claim would imply a theory of growth largely driven by geography and climate. For if the recent increase in temperature in Africa pushed the continent further away from the optimal temperature for economic activity (which, according to the authors’ model, is around 13 degrees Celsius), then the fact that Africa was, even before anyone heard of climate change, warmer that this optimal temperature, must have historically had negative effect on Africa’s growth.
We are thus facing here a variant of economic growth theories that put the emphasis not only on exogenous factors, not only on geography (like navigable rivers, impassable mountains) but on specific exogenous geographic factors like climate. The growth regression (the step No. 3) reported in the paper is breathtaking in its simplicity. It is a country fixed effect panel regression where country’s growth rate depends on contemporaneous temperature and precipitatons (both linearly and squared), country and time fixed effects…and nothing else! No employment, no capital, no saving rates, no institutions, no civil wars…
I will leave it at that: the climatological explanations have been used for many things: from Montesquieu who thought that climate explained differences in political systems to Paul Bairoch’s argument regarding the non-transmission of the agricultural revolution.
But let’s suppose that this explanation is true and that indeed, as the authors claim, climate change was responsible for slowing the growth of poor countries. That would have enormous consequences (which, by the way, they, at least in this paper, fail to mention).
Since the change in climate is brought about by historical emissions of the currently rich counties (the stock effect), and by their current emissions plus those of China (the flow effect), this means that the very growth in the North is directly responsible for the lack of growth in the South. The implication is quite extraordinary. In the past, dependencia theorists proposed that the “center”, the Global North, deepened the underdevelopment of the Global South through a division of labor that let the South produce only agricultural goods; or that the Global North helped develop only some parts of Southern economies while leaving the rest underdeveloped. Such theories saw delinking from the North as a solution.
But the important point is to notice that in those theories the integration of the Global North and the Global South was bad for the South; in the new “climate theories” it is simply the fact that the North grows that is bad. It need not interact with the South of all. Northern growth alone makes the South poorer. This is quite extraordinary. It is not that my exploiting or cheating somebody is a condition for my wealth; it is that my wealth as such (acquiring it with no interaction with the injured party) is a bad news for somebody else (Africa in this particular case).
Moreover, it means that the very growth of the Global North makes the reduction or the eventual elimination of African poverty difficult, if not impossible. If we are to believe the authors, then every percentage point of additional GDP in the North makes conditions in Africa worse and the reduction of poverty there more difficult.
Thus for the elimination of global poverty, we need drastic reduction in emissions which means absolute reduction of Northern incomes, namely, a steady negative rate of growth of rich countries.
I will leave it to the reader to reflect on how politically feasible such a solution is (I wrote about that before, here and here)—but I think that the enormity of the implications of these results should be realized. Now, whether the results really make sense or not, whether the level of temperature by itself is a significant explanatory factor of economic development is a thing that needs to be proved. More panel regressions of economic growth? I thought we left them behind in the 1990s but perhaps I was wrong.