Masayuki Kudamatsu
Osaka School of International Public Policy (OSIPP)
Osaka University
16 May, 2018
Natural science data #1
Natural science data #2
(The above image refers to the mean temperature for 1951-1980)
Natural science data #3
1
2
A cause of ethnic diversity
Impact of Tsetse flies
on African economic development
3
Origin of states
Before diving in to these 3 examples...
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
Natural science data #1
This data is created as follows:
[1] Obtain spatial distribution of croplands from satellite images
[2] Divide the earth into 0.5-degree cells
[3] Calculate % of cells cultivated
[4] Regress [3] on degree-days, moisture, soil carbon density, soil pH
[5] Predict % of cells cultivated from [4]
This data is created as follows:
[1] Obtain spatial distribution of croplands from satellite images
[2] Divide the earth into 0.5-degree cells
[3] Calculate % of cells cultivated
[4] Regress [3] on degree-days, moisture, soil carbon density, soil pH
[5] Predict % of cells cultivated from [4]
e.g. wars, population growth
In economics, it's been known that
ethnic diversity is associated with lower economic growth
In antholopology, some studies suggest
linguistic diversity is associated with ecological diversity
Stelios Michalopoulos,
an economist at Brown University,
provides systematic evidence
for language-ecology relationship
in his 2012 paper.
A seemingly irrelevant topic in social science:
Calculate the standard deviation of suitability for agriculture
within each country
Count the number of languages spoken within each country
from the World Language Mapping System
Source: Figure 2 and page 1514 of Michalopoulos (2012)
# of languages
Nepal
Greece
Source: Figure 4 of Michalopoulos (2012)
Japan
Source: Figure 4 of Michalopoulos (2012)
Saudi Arabia
Source: Figure 4 of Michalopoulos (2012)
Ethiopia
Source: Figure 4 of Michalopoulos (2012)
Senegal
Source: Figure 4 of Michalopoulos (2012)
Source: Figure 4 of Michalopoulos (2012)
Swaziland
Source: Figure 4 of Michalopoulos (2012)
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
Source: Figure 4 of Michalopoulos (2012)
Source: Map 1 of Online Appendix to Michalopoulos (2012)
Natural science data #2
(The above image refers to the mean temperature for 1951-1980)
Create the start-of-the-art model of climate
at the spatial resolution of 2-degree cells across the earth
"Forecast" the weather 6 hours ago from the current weather
Correct the "forecast" whenever actual observations are available
Repeat this up to the year of 1871
A seemingly irrelevant topic in social science:
Image sources: Encyclopedia Britannica and Wikipedia
Manure
Draft power
Anthropologists and historians speculate
Tsetse flies were the main cause for
Africa's economic backwardness
Africa is historically...
agricultural productivity: low
population density: low
Marcella Alsan
health economist in Stanford University
provides systematic evidence
on the impact of tsetse flies in her 2015 paper
Source: Figure 2 Panel A of Alsan (2015)
Source: Figure 2 Panel B of Alsan (2015)
Temperature & humidity
in 1871
Source: Figure 3 Panel A of Alsan (2015) and Wikipedia
Source: Figure 5 Panel A of Alsan (2015)
Tsetse Suitability Index
Ethnographic Atlas
Source: Figure 4 Panel A of Alsan (2015)
Source: Figure 4 Panel A of Alsan (2015)
Source: Figure 4 Panel A of Alsan (2015)
Source: Figure 4 Panel A of Alsan (2015)
(Over 20,000)
Source: Figure 4 Panel A of Alsan (2015)
Source: Figure 4 Panel A of Alsan (2015)
Natural science data #3
For 5 arc-minute cells across the earth
Daily average weather
temperature, precipitation, humidity, wind speed, sun exposure
Soil type, Elevation, Land gradient
For each crop (wheat, rice, potatoes, cassava, etc.)
Crop growth parameters
How sensitive to degree-days, water scarcity, etc.
at each of the four stages of crop growth
A seemingly irrelevant topic in social science:
Historians and anthlopologists debate
whether agriculture led to the emergence of states
A seemingly irrelevant topic in social science:
These four economists argue
it's appropriability of crop harvests that matters for states to emerge
Luigi
Pascali
(Pompeu Fabra)
Omer
Moav
(Warwick)
Zvika
Neeman
(Tel Aviv)
Joram
Mayshar
(Hebrew)
Wheat, rice, maize...
Storable
Harvest within a short season
Cassava, yam, taro, bananas...
Perishable upon harvest
Harvesting is non-seasonal
Convert each crop's
potential yield
into calory units
Step 1
Obtain the maximum caloric yield among cereals
Step 2
Obtain the maximum caloric yield among roots and tubers
Step 3
Step 4
Take the difference
Source: Figure 7 of Mayshar, Moav, Neeman, & Pascali (2015)
Source: Figure 4 of Mayshar, Moav, Neeman, & Pascali (2015)
Use Ethnographic Atlas across the world
on average
1.89
2.13
Relative
productivity
of cereals
higher
by 1 s.d.
1.65
Relative
productivity
of cereals
lower
by 1 s.d.
Data science's most basic tool:
Image adapted from Figure 3.1 of James et al. (2013)
Y
X
Image adapted from Figure 3.1 of James et al. (2013)
Y
X
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
Image adapted from Figure 3.1 of James et al. (2013)
X
Y
This lecture is based on the following academic articles:
Mayshar, Joram, Omer Moav, Zvika Neeman, and Luigi Pascali. 2015.
"Cereals, Appropriability and Hierarchy."
CEPR Discussion Paper, no.10742.
Alsan, Marcella. 2015.
"The Effect of the TseTse Fly on African Development."
American Economic Review, 105(1): 382–410.
Michalopoulos, Stelios. 2012.
"The Origins of Ethnolinguistic Diversity."
American Economic Review, 102(4): 1508–1539.
Python
CodeCademy: Learn Python (for free)
R
DataCamp: Introduction to R (for free)
The Owen Guide (for free)