On the Spatial Distribution of Development. The Roles of Nature and History.

31st Jan 2017

Authors: Vernon Henderson, Adam Storeygard, Tim Squires, David Weil

What determines where people live? Why are some places so densely populated and some so empty?  In daily life, we take this variation in density as a matter of course, but in many ways it can be quite puzzling. 

Worldwide population density


Economists point to three factors to explain how population is distributed. The first is that there are differences in geographical characteristics, often referred to as “first nature,” that make some places more amenable for living or producing output than others.  This explains why mountainous regions, deserts, tundra and so on tend to have low population density, and why much of the world’s population is situated in places where it is relatively easy to produce food. 

The second factor is agglomeration. Because of economies of scale and gains from trade, we humans often find it efficient to gather in small areas.  Of course, many industries, most notably food production, don’t benefit from such concentration, and are instead spread out in accord with first nature resources.  Further, there are limits to the benefits of agglomeration: because of congestion and transport costs, the urban population is spread among many cities, which are in turn spatially dispersed.  

The final factor is history: cities, once established, have a very strong tendency to stay put.  This persistence results from many factors, often collectively described as “second nature” (Cronon, 1992).  Among these factors are long-lived capital, political power, and the fact that once agglomeration has started in a particular place, it will be a natural focus for future equilibria.  As discussed by Bleakley and Lin (2012) and Michaels and Rauch (forthcoming), this persistence can be important even when the reasons that a city has been established in a particular place are no longer important. 

In “The Global Spatial Distribution of Economic Activity: Nature, History, and the Role of Trade,” we ask how certain economic and technological developments have changed the ways in which “first nature” characteristics impact the population distribution. Simple examples of such developments are the impacts of air conditioning, irrigation, and the discovery of new uses for particular mineral resources. We focus on two natural characteristics where we think economic and technological developments have been most important; these are the suitability of a region for growing food, and the suitability of a region for engaging in national and international trade. 

Over the last several centuries, the importance fertile land as a determinant of population density has declined both as agricultural productivity has increased – releasing labor from farms - and transport costs have fallen – so people don’t need to locate where food is produced.  Lower transport costs, along with increased opportunities for gains from trade, have similarly raised the value of locations, such as those on coasts, navigable rivers, or natural harbors, that are accessible to trade, either within or between countries. 

Our goal is to show how these historical changes are reflected in the distribution of population today, but before looking at the role of history, we start by simply examining the explanatory power of first nature characteristics for population distribution in the world at present.

 

First Nature Data and the Distribution of Population Today

Our starting point in measuring the dispersion of population is lights observed at night by weather satellites. Specifically, we use the 2010 Global Radiance Calibrated Nighttime Lights dataset (Ziskin et al. 2010). In previous work (Henderson, Storeygard, and Weil, 2012), we showed that change over time in night-lights data is a useful proxy for the growth of GDP in countries with poor national income accounts data.  The lights data are distributed as a grid of pixels of dimension 0.5 arc-minute resolution (1/120 of a degree of longitude/latitude). We aggregate into a grid of 1/4-degree squares, with each square covering approximately 770 square kilometers at the equator.  At this resolution our sample is roughly 240,000 grid squares (excluding squares made up solely of water).  Figure 1 shows this grid cell data for the world as a whole.    

 

fig1 spatial

 

The first-nature variables we use in predicting lights come in three groups.  The first, “agriculture”, are factors that seem clearly related to producing food.  These comprise six continuous variables (temperature, precipitation, length of growing period, land suitability for agriculture, elevation, and latitude) as well as a set of 14 indicators for biomes (mutually exclusive regions encoding the dominant natural vegetation expected in an area, based on research by biologists.)  The second group, “trade”, focus on access to water transport. These are dummy variables for whether the center of a grid cell is within 25 kilometers of a coast, navigable river, major lake, or natural harbor, as well as a continuous measure of distance to the coast.  Finally, we define a “base” group of two variables -- ruggedness and malaria ecology -- which seemed to us to be roughly equally relevant for agriculture and trade. 

Figure 2a shows the fitted values from regressing lights in a grid square on our first nature variables.  Together, these variables explain 47% of the variation in lights, and looking at the figure, there is clearly a strong resemblance between the fitted values and the world as we know it.  However, there are two issues with jumping from this result to the conclusion that nature really does explain such a large fraction of variability in population density.  The first is that variation in visible light in not solely determined by population density.  The other big determinant is income per capita. This explains why Japan is so much brighter than Bangladesh in Figure 1 even though the latter is more densely populated.  The second problem is that a statistical correlation between geographical characteristics and either income or population density might not indicate a true effect of geography, but rather be proxying for the effect of something correlated with geography such as institutions (Acemoglu et al, 2001).

 

fig2 spatial

 

Both of these problems are addressed by looking at variation in lights and natural characteristics within countries. Formally, this amounts to including country fixed effects in our analysis, which is done in all the work reported below. Figure 2b  now shows the effect of first nature characteristics using only within country variation in lights. As the figure shows, knowing only how geography affects population within countries, one would still do a pretty good job of predicting the variation in population density the world over.  

 

fig3 spatial

 

  The Changing Importance of First-Nature Characteristics

We now turn to the question of how the importance of natural characteristics as a determinant of the distribution of population has changed over time. In pursuing this goal, the effect of persistence, as described above, turns out to be very important.  We are interested in how technology in historical times affected agglomeration at those times, but the data on population density that we use (described below) is only available for the world today.  However, if we know when (in a rough sense) agglomeration began in a country, then we can use the similarity of today’s distribution to the historical distribution to learn about how the technology available at that time affected agglomeration.   

A key to our approach is comparing countries where agglomeration took place early, thus reflecting the weights put on natural characteristics further back in time, with those that agglomerated later.  Unfortunately, we don’t have a good, consistent measure of exactly when agglomeration took place, so instead we use data from 1950, on both urbanization as well as two proxies: education and GDP per capita.  Our assumption is that countries with higher values of these measures as of that point in time also started their urbanization process earlier.

We use several statistical approaches to parse the data. One is to estimate coefficients on our “agriculture” and “trade” variables separately for early and late agglomerators, while simultaneously letting the data determine where the cutoff is between these two groups of countries (essentially, looping through all possible division points to find the one that give the best fit).  Applying this method using urbanization in 1950, for example, we find that the cutoff between early and late agglomerators is an urbanization rate of 36.2%, which puts 70 out of 189 countries (57.2% of our grid squares) in the “early” category.

Table 1 then shows the R-squareds from running regressions of visible lights on either the set of base variables (including country fixed effects), the base plus agriculture variables, or the base plus trade variables.  The improvement in fit that comes from adding agricultural variables is much larger in the early than in the late agglomerating countries; correspondingly, the improvement in fit that comes from adding trade variables is much larger in the late agglomerating countries than in the early agglomerators. We find similar patterns when we use education or GDP per capita in 1950 to split the data, and also when we look solely within the New World or the Old World.

 

Table 1: R-squareds
RHS variables Early Agglomerators Late Agglomerators
Base .350 0.359
Agriculture + Base 0.613 0.508
Trade + Base 0.385 0.447

 

These results tell what at first seems to be a puzzling story: Late agglomerators are generally poorer countries, and on average are more dependent on agriculture than early agglomerators.  Yet it is the early group of countries in which agricultural variables do a better job of predicting the location of population and economic activity. Our explanation of this apparent puzzle looks to the timing of when agricultural productivity rose and similarly when trade costs fell.  In countries where agglomeration got going early, the rise in agricultural productivity preceded the decline in transport costs, as shown in Figure 3.  That is, people began moving from farms to cities as a time when it was still relatively expensive to move food from place to place.  As a result, cities were located close to areas conducive to food production.  By contrast, in late agglomerators, the rise in agricultural productivity that allowed urbanization came later relative to declining transport costs, and so the latter was relatively more influential as a determinant of location

 

Figure 3

 fig4 spatial

 

An interesting implication of this analysis is that countries that are only urbanizing now have population distributions that are more appropriate to modern technology than do those that urbanized earlier.  For example, even though in Europe coastal areas already have particularly high population densities, our estimates imply that if Europe had developed later, coastal density would be even greater.  Similarly, if Africa had developed earlier, interior areas such as the Ethiopian highlands and the Congo basin would have higher relative population densities than what is actually observed today.  

Another set of implications from our paper related to spatial inequality within countries. We expect early agglomerators to have a higher degree of spatial equality in lights than late agglomerators as their activity is focused on agriculturally suitable land, and their population distribution was inherited from a period when transport costs were high.  To assess this prediction, we calculate a spatial Gini coefficient for light across grid cells for each country.  Table 2 shows the results from regressing the lights Gini on urbanization in 1950.  The coefficient is negative, as predicted.  Further, controlling for the Gini of lights predicted using our geographic variables, as well as measures of country size and population (and thus population density) does not affect the result.  

 

Table 2:  Gini Coefficient of Lights
Urbanization in 1950

-0.00400

(0.00067)

-0.00425

(0.00063)

-0.00285

(0.00049)

Gini Coefficient of Predicted Lights  

 0.382

(0.073)

 0.0933

(0.0658)

ln(land area)    

 0.0864

(0.0074)

ln(population in 2010)    

-0.0500

(0.0081)

Constant

 0.851

(0.026)

 0.702

(0.037)

 0.212

(0.063)

Observations 131 131 131
R-squared 0.167 0.277 0.597

 

Conclusion

The saying that “geography is destiny” is often attributed to Napoleon.  Meanwhile, the American industrialist Henry Ford really did say that “history is bunk.”  In our research, we have shown that, when it comes thinking about how population is distributed within countries, there is reason to doubt both of these statements.  Geography clearly matters quite a bit, but the aspects of geography that matter change over time.  Further, there is enormous persistence in location, so that the ways in which geography mattered historically are still reflected in the spatial distribution of population today. 

To many readers, sitting in cities founded hundreds of years ago, sipping coffee grown thousands of kilometers away, none of this will come as a great surprise.  But understanding the dynamic interplay of geography, technology, economic growth, and history -- a project in which our paper is step -- is of great importance in thinking about many issues facing the world today.  Not only are the impacts of different geographic characteristics continuing to change with economic and technological development, but in decades to come, geographic characteristics themselves will be changing at an ever increasing rate.  At the same time, the locational decisions made today will have impacts in centuries to come. 


References

Acemoglu, Daron, Simon Johnson ,and James  A. Robinson. "The Colonial Origins of Comparative Development: An Empirical Investigation." The American Economic Review 91.5 (2001): 1369-1401.

Bairoch, Paul (1988), Cities and Economic Development, Chicago: University of Chicago Press.

Bleakley, Hoyt, and Jeffrey Lin. "Portage and path dependence." The quarterly journal of economics 127.2 (2012): 587.

Cronon, William. Nature's metropolis: Chicago and the Great West. WW Norton & Company, 1992.

Henderson, J. Vernon, Adam Storeygard, and David N. Weil. 2012. “Measuring Economic Growth from Outer Space.” American Economic Review 102(2): 994-­1028

Michaels, Guy, and Ferdinand Rauch. "Resetting the urban network: 117-2012." The Economic Journal, forthcoming.

Mohammed, Saif I. Shah, and Jeffrey G. Williamson. "Freight rates and productivity gains in British tramp shipping 1869–1950." Explorations in Economic History 41.2 (2004): 172-203.

Ziskin, Daniel, Kimberly Baugh, Feng Chi Hsu, Tilottama Ghosh, Chris Elvidge. 2010. “Methods Used For the 2006 Radiance Lights.” Proceedings of the 30th Asia­Pacific Advanced Network Meeting 131­-142.

 

A longer version of this article was originally published in The Long Economic and Political Shadow of History, Vol 1

To see the full length journal, please click here