The Ghosts of Climates Past and the Specters of Climates Future
3. The Impact of Climates Present
Modern views of economic development give short shrift to climate as the basis for the differences of the wealth of
nation^.^.
A review of a hand- ful of textbooks on economic development shows that climate is confined to a few lines in hundreds of pages.3 The modern view of economic growth presents development as a vehicle driven by the four wheels of capital, labor, resources, and technology, and only with a major stretch of interpretation would we equate resources with climate. The recent wave of studies of inter- national differences in productivity have never included climate as a deter- mining variable. Modern trade theories have had difficulty finding empirical evidence of the power of Heckscher-Ohlin trade theories that emphasize that international trade is based on a region's resource base, which base would include climate along with other factor endowments.2A notable exception is a study by Kamarck (1976). This essay attempts to explain why it is that countries in the tropical regions are so burdened by laterite soils, weeds, locusts, parasites, and heat stress that, "all other relevant factors being equal, the pact of development in tropical countries tends to be slowern (Kamarck, 1976, p. 4)
3 F ~ r example, see Streeten (1972), Nafiiger (1990), Todaro (1991). Perhaps the most striking irony is the treatment by Todaro. He states, "It is a historical fact that almost every successful ezample of modern economic growth has occurred in a temperate-zone country." (Italics in original) He then takes nine lines of a large tome, with not a single reference, to speculate on this remarkable "fact".
How can we rationalize modern attitudes toward climate? One possible reason for ignoring the importance of climate is that there seems little point in studying the influence of so exogenous a variable. A more serious possibil- ity is t h a t human societies can adapt t o whatever climate is dealt by nature.
We explore briefly in this section this view and then examine some economic and geophysical d a t a t o determine whether they support the earlier climatic determinism or the alternative adaptive view.
In opposition t o Huntington's climatic hypothesis of civilization would be what might be called the adaptive theory of climatic impact. The adap- tive theory holds that in the very long run, humans are essentially nomadic toolmakers. Homo adaptus thrives in every climate from Hong Kong t o Helsinki: with a few exceptions of diminishing importance, humans can in- vent products or processes that can offset the disadvantages of climate, or can by trade turn apparent disadvantage t o economic profit. At the extremes of cold, Sherpas make a relatively good living leading people up Mt. Ever- est, while deserts, free of dreaded ice and snow, are increasingly attractive places t o old folks as places t o putt and putter during their retirement. And in the end, if some region is so barren as to attract no conventional economic activity, it will probably become attractive as a wilderness retreat.
In the short run, of course, life can be full of surprises and dreadful climatic shocks. Homo adaptus will still be rendered homeless if hurricanes destroy flimsy structures, if storm surges break across land that is not meant for risk-free habitation, if farmers cultivate drought-prone land, or if civil wars destroy the land or dampen entrepreneurial impulses. To say that humans are adaptive says nothing about whether the adapted standards of living are high or low, whether nations are a t war or peace, whether the climatic shock is the straw that breaks the camel's back, or whether the appropriate policies are t o prevent the laying of that last straw through massive investments t o slow climate change rather than t o lighten the load from other burdens so that occasional climate shocks are nuisances rather than catastrophes and therefore break nothing.
Figure 2 illustrates the adaptive view of climatic impacts. The horizon- tal axis represents a synoptic climate variable such as average temperature, while we measure a region's productivity on the vertical axis. The horizon- tal line LR represents the long-run productivity of a region and suggests t h a t productivity is independent of the climate. In the short run, however, productivity will be maximized a t the "design climate" - we therefore repre- sent SRo as the productivity curve that corresponds t o capital, management, infrastructure, and localized technologies that are designed for climate To.
If climate were t o change t o T I , cool-weather crops would wilt, ski areas
To T I
would fail, and other signs of a mal-adapted technology would emerge, with the equilibrium moving from A to B and productivity falling from P* t o P I . Over time, however, the economy would adapt as tropical fruits replace temperate grains and campers replace skiers. Once all the adaptations had taken place, productivity would rise t o point C, with a productivity equal t o the initial level and with a new short-run productivity curve of SR1.
Figure 2 makes two other important points about adaptive systems. The first and not surprising corollary is that large shocks have a bigger short- run impact than do small shocks. The second and surprising implication is that, t o a first-order of approximation, small shocks have no impact; more precisely, a small climate shock in the short run, before any adaptation has taken place, has the same impact as a long-run shock, or one after adaptation has taken place. In terms of Figure 2, a small shock will move along the short-run productivity curve, which is tangent t o the long-run productivity curve a t the initial point.
There are many potential objections t o the adaptive view shown in Fig- ure 2. One valid and enduring objection is simply that some aspects of social or natural systems cannot adapt or adapt so slowly that significant damages cannot be avoided. Forest ecosystems, coral reefs, cultural treasures like
----
-./-z: - C-
----- .
LRVenice, wildlife refuges like Yellowstone - for these, the possibilities of adap- tation seem dubious, so the short-run curve in Figure 2 might well be for 2 centuries rather than 2 years. A second class of objection is that the long- run curve is not horizontal but would show marked losses as climate changes.
Such a case might arise if climate change involves global glaciation, if there is a shift in ocean currents that changes the climate of the North Atlantic communities into that of Alaska, or if midcontinental warming and drying destroys the globe's grain belts.
Using modern income concepts, we might examine very simply the rela- tionship between certain geophysical variables and different economic mea- sures. For geophysical d a t a t o represent climate, we use two very simple vari- ables, latitude and average temperature. The latitude variable, measured as the latitude of the geographical center of the country for 77 countries, is a crude way of representing average temperature and other variables related t o solar intensity. The temperature variable is the average of the July and January maxima and minima for forty countries. Clearly these variables are too aggregated for large countries like the United States, which spans 52O of latitude from Hawaii t o Barrow, Alaska. But for smaller countries like Belgium or Gabon, the climatic variables not mask great diversity.
For economic variables, we begin with 1987 per capita GNP, using the purchasing-power parity corrected exchange rates developed by the interna- tional income comparison project. We have adjusted these t o reflect more realistic d a t a on the per capita incomes of the formerly centrally planned economies. Figures 3 and 4 show the relationship between the geophysical variables and per capita incomes. The circles on the graphs are of three dif- ferent shapes. The largest circles are for countries with 1987 GNPs of more than $100 billion; the medium circles are for those countries with 1987 GNPs of between $20 and $100 billion; and the smallest circles are for countries with 1987 US GNPs of less than $20 billion.
Figure 3 indicates that the zone from lo0 South t o 20° North is an economic desert, with virtually no countries that have achieved high levels of per capita income. However, in the latitudes poleward from 35O North or 25O South, there is virtually no relationship between latitude and economic performance.
Figure 4 shows the relationship between per capita income and temper- ature for a smaller sample of countries. Over most of the range, from a mean temperature of 40°F t o around 65OF, there is no relationship between mean temperature and income per capita. Above 65OF average temperature there is only a handful of countries, but it is interesting t o note that no countries
-40 -20 0 20 40 60 80
Latitude
Figure 3. Geography and income (latitude and income per capita).
in this sample with a mean temperature of over 65OF has risen above the
$2000 per capita income level.
The measure of income per capita is defective in that human migrations will over a scale of centuries tend to equalize incomes in different regions (although the tendency is clearly quite weak in light of current findings on the lack of convergence among countries). In 1980, Alaska has the highest per capita income of any of the 50 United States, and we might therefore mistakenly conclude that the state of Alaska has a very fertile climate. A better measure of the economic clemency of a climate would probably be the "Ricardian rent" that land in any area yielded. Failing that, we could examine the income per unit area, measured as a country's GNP per unit area (in 1987 US dollars per square kilometer).
Figures 5 and 6 show the results of this calculation. Figure 5 shows the same latitudinal results as Figure 3 - that there is an economic desert in the lowest latitudes. But the interesting new feature here is t o show that the economic return per unit land peaks in the middle latitudes - say between 40 and 50 North and around 30 South - and then seems t o decline again a t the highest latitudes. This result reflects the sensible result that the highest latitudes (Alaska, Northern Canada, Siberia) may have high incomes per person but few persons are able to earn a high income there.
60 70
Temperature
Figure 4. Geography and income (temperature and income per capita).
-40 -20 0 20 40 60 80
Latitude
Figure 5. Geography and income (latitude and income per area).
40 50 60 70 80 90
Average temperature
Figure 6. Geography and income (temperature and income per area).
Figure 6 seems more of a standoff between the variables. Income per unit area is highest in temperature ranges that are more moderate, and there is a modest hump-shaped relationship. But no clear threshold appears in these data.
A final word should be added to put these geophysical variables in a larger context. Climate may have an effect upon incomes, but on the whole the effect is swamped by other variables. Looking a t Figure 5, we see that incomes per square kilometer vary from a low of about $31 for China to a high of about $30,000 for Hong Kong, or from $37 per kilometer in Indonesia t o $6200 in Japan. Latitude explains less than 1% of the variance in income per capita or per area. We should surely look t o factors other than climate t o explain most differences in the wealth or poverty of nations.