Climate, Geography, and the Origin of Economic and Political

Climate, Geography, and the Origin of Economic and Poli8cal Ins8tu8ons Roy Elis and Stephen Haber Stanford University Presented at New Zealand Ministry of the Treasury December 3, 2015 Why do income and democracy cluster geographically? Why do income and democracy cluster geographically? What do we claim? 1.  Climate and geography explain 1/3 to 1/2 of the variance of GDP/c and democracy across the world. 2.  Climate and geography work on GDP/c and regime types work indirectly, by influencing fundamental ins8tu8ons. 3.  Fundamental ins8tu8ons emerged out of the strategies that socie8es could pursue to solve the problem of food scarcity and uncertainty at the beginning of the modern era. What is the chain of reasoning that supports our claims? How,Geography,and,Climate,Influence,LongMRun,Paths,of,Development
Geography
State,of,Technology,Circa,1800
Climate
Capacity,to,Grow,,Store,,&,Transport,Food
Frequency,of,,Weather,Shocks
Incentives,to,Invest,in,Human
,Capital,and,Property,Rights
Institutions,in,the,Past
Type,of,State,that,Emerged:
Transactional;,Insurance;,or,Fragile
,
,
Level,of,
Democracy,
Today
Level,of,
GDP/c
Today
The origin and outcome of Transac'onal States Low transport costs + climate suited to genera8on of storable agricultural surpluses + idiosyncra8c shocks à Strong incen8ves to mi8gate scarcity and uncertainty through local and regional tradeà Strong incen8ves to mi8gate the problem of theWà * Strong incen8ves for agents to invest in human capital * Strong incen8ves to invest in ins8tu8ons to enforce property/contract rights * Strong incen8ves to limit the discre8on of the government à high popula8on density à high level of economic development in the long run; à high probability of consolida8ng democracy. How do storable crops and low transport costs give rise to Transac8onal States? Each farmer is isolated, and faces scarcity and uncertainty on his own
Problems of scarcity and uncertainty caused by idiosyncra8c shocks can be solved by local trade, provided that ins8tu8ons protect transac8ons Property rights institutions facilitate transactions among agents
Trading networks encourage specializa8on—
which is human capital Property rights institutions creates incentives to specialize,
and increase the range of transactions
In 8me the transac8ons network becomes so dense that its origins are difficult to see Which gives rise to a transaction society, in which the original agents
transacting are not only a dim memory
What would happen if farmers could not use local and regional trade to mi8gate scarcity and uncertainty, because weather shocks were temporarily and spa8ally correlated? The Insurance State
Agricultural
Producers
State
The origin and outcome of Insurance States Low transport costs + climate suited to genera8on of storable agricultural surpluses + frequent aggregate weather shocksà Strong incen8ves to mi8gate scarcity and uncertainty through centralized storageà * Weak incen8ves for agents to invest in human capital * Strong incen8ves to create a centralized government capable of taxing, storing, and protec8ng surpluses * Weak incen8ves to limit discre8on of government à high popula8on density à middle level of economic development in the long run; à low probability of consolida8ng democracy. The origin and outcome of Fragile States High transport costs or climate not suited to genera8on of storable agricultural surplusesà Difficult to trade, nothing to store centrallyà * Weak incen8ves for agents to invest in human capital * Property rights enforced through disorganized violence * Weak incen8ves to limit discre8on of government à low popula8on density à low level of economic development in the long run; à low probability of consolida8ng democracy. How do we develop evidence in support of our chain of reasoning? •  We tried to go back 400 years to randomly assign countries to treatment and control groups •  Because that did not work, we developed geo-­‐coded (GIS datasets) for the en8re world on navigable waterways before steam power, natural harbors, terrain slopes and eleva8ons, the frequency of aggregate weather shocks, soil suitability for growing storable crops, malaria ecology, and climate suitability for storing those crops. •  We also take advantage of natural experiments of history and conduct compara8ve historical case studies. What is our unit of analysis, and how do we measure it? NYC 95,461 km2 Mexico City 12,666 km2 Kinshasa 152,757 km2 We es8mate each hinterland’s poten8al produc8on of cereals using data on soil characteris8cs from FAO GAEZ3 Kcal%poten+al%produc+on%(in%billions)%
60,000$
50,000$
Kcal%poten+al%in%billions%
40,000$
30,000$
20,000$
10,000$
0$
Series1$
New$York$
Mexico$City$
Kinshasa$
52,156$
4,032$
21,855$
We next es8mate the rela8ve storability of those cereals as a func8on of temperature and humidity We can then es8mate the storable kilocalorie poten8al of each hinterland Storability-­‐adjusted, cereal kilocalorie poten8als cluster geographically Storability-­‐adjusted, cereal kilocalorie poten8als cluster geographically We can then es8mate the prevalence of malaria at the hinterland level, which affected storable kilocalorie produc8on by reducing labor effort We then use data from the NOAA to es8mate the frequency of aggregate weather shocks for each hinterland since the 19th century Aggregate weather shocks also cluster geographically Aggregate weather shocks also cluster geographically Is there a rela8onship between current levels of democracy and the ability to grow and store calories and aggregate weather shocks around hinterland of the largest city in 1800? Storable
Cereals,
Weatherthe Shocks,
and Average
Polity
(1975−2014)
High Malaria
Low Malaria
Weather shocks
(log percent months PDSI above 4.0)
3
2
Polity Index (1975−2014)
Democracies [86.9, 100]
Anocracies [26.9, 86.9)
1
Autocracies [0, 26.9)
0
−1
9
12
15
18
9
Storable Cereals
(log kcal indexed by UK storability)
12
15
18
Is there a rela8onship between levels of GDPc in 2000 and the ability to grow and store cereals and aggregate weather shocks around he hinterland of tGDP
he largest city in 1800? Storable
Cereals,tWeather
Shocks, and
Per Capita
(2010)
High Malaria
Low Malaria
Weather shocks
(log percent months PDSI above 4.0)
3
2
Log per capita GDP (2010)
High [10.1, 11]
Middle [7.7, 10.1)
1
Low [0, 7.7)
0
−1
4
8
12
16
4
8
Storable Cereals
(log kcal indexed by UK storability)
12
16
How do we know that climate and geography work on GDP and Democracy through the level and of human capital n the past? Storabledistribu8on Cereals, Weather Shocks,
and Human
Capital i(1910)
High Malaria
Low Malaria
Weather shocks
(log percent months PDSI above 4.0)
3
2
Numeracy (1910)
Highly Numerate
(Whipple below 110)
Moderately Numerate
(Whipple between 110 and 175)
Innumerate
(Whipple at or above 175)
1
0
−1
4
8
12
16
4
8
Storable Cereals
(log kcal indexed by UK storability)
12
16
How do we know that there is a rela8onship between human capital in the past and GDP/c today? And how do we know that there is a rela8onship between human capital in the past and democracy today? Links in our chain of reasoning that s8ll require evidence Measuring the ins8tu8ons that protected transac8ons (the frequency of urban popula8ons >10,000 in 1800—under construc8on). Do our claims stand up to alterna8ve reasons? The panern of democracies and autocracies across former Bri8sh colonies Do our claims stand up to alterna8ve reasons? Do our claims stand up to alterna8ve reasons? Do our claims stand up to alterna8ve reasons? Do our claims stand up to alterna8ve evidence? The quasi-­‐natural experiment of the former USSR Do our claims stand up to alterna8ve evidence? Natural experiments from history 1. 
2. 
3. 
4. 
Hawaii Australia The Colonial United States Providence Island and Massachusens Bay Do our claims stand up to out of sample tests? 1.  The distribu8on of democracy/economic development in an8quity: Greece, Rome, Mesopotamia, Egypt, and China. 2.  The distribu8on of poli8cal ins8tu8ons and economic development across Na8ve American cultures prior to 1492. Work to be done 1.  Moving beyond the hinterland of the largest city in 1800 to an endogenous network of hinterlands. 2.  Learning a whole lot more history…. Extra slides What happens if we drop the two strange cases (North Korea and Swaziland)? The rela8onship between rainfall and democracy The rela8onship between rainfall and the ability to grow and store cereals