NBER WORKING PAPER SERIES
CLIMATE CHANGE AND ECONOMIC GROWTH:
EVIDENCE FROM THE LAST HALF CENTURY
Benjamin F. Jones
Benjamin A. Olken
Working Paper 14132
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
We thank Daron Acemoglu, Esther Duflo, Michael Greenstone, Jonathan Gruber, Seema Jayachandran,
Charles Jones, Peter Klenow, William Nordhaus, Elias Papaioannou, and Carl Wunsch for helpful
comments and suggestions. The views expressed herein are those of the author(s) and do not necessarily
reflect the views of the National Bureau of Economic Research.
NBER working papers are circulated for discussion and comment purposes. They have not been peerreviewed or been subject to the review by the NBER Board of Directors that accompanies official
© 2008 by Melissa Dell, Benjamin F. Jones, and Benjamin A. Olken. All rights reserved. Short sections
of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full
credit, including © notice, is given to the source.
Climate Change and Economic Growth: Evidence from the Last Half Century
Melissa Dell, Benjamin F. Jones, and Benjamin A. Olken
NBER Working Paper No. 14132
JEL No. O11,O13,O40,Q54
This paper uses annual variation in temperature and precipitation over the past 50 years to examine
the impact of climatic changes on economic activity throughout the world. We find three primary results.
First, higher temperatures substantially reduce economic growth in poor countries but have little effect
in rich countries. Second, higher temperatures appear to reduce growth rates in poor countries, rather
than just the level of output. Third, higher temperatures have wide-ranging effects in poor nations,
reducing agricultural output, industrial output, and aggregate investment, and increasing political instability.
Analysis of decade or longer climate shifts also shows substantial negative effects on growth in poor
countries. Should future impacts of climate change mirror these historical effects, the negative impact
on poor countries may be substantial.
Department of Economics
Massachusetts Institute of Technology
50 Memorial Drive, Cambridge MA 02142
Benjamin F. Jones
Kellogg School of Management
Department of Management and Strategy
2001 Sheridan Road
Evanston, IL 60208
Benjamin A. Olken
Harvard Society of Fellows
78 Mt. Auburn St.
Cambridge, MA 02138
Climate change may – or may not – be a central issue for the world economy. Yet
assessing the economic impact of climate change faces a fundamental challenge of complexity:
the set of mechanisms through which climate may influence economic outcomes, positively or
negatively, is extremely large and difficult to investigate comprehensively. Even if the effect of
climate on each relevant mechanism were known, one would still be faced with the challenge of
how various mechanisms interact to shape macroeconomic outcomes.
The complexity of the climate-economy relationship is apparent in a brief survey of the
literature. Much research focuses on agriculture (e.g., Adams et al. 1990; Mendelsohn et al.
2001; Deschenes and Greenstone 2007; Guiteras 2007). Other research examines ocean fisheries,
fresh water access, storm frequency, migration, tourism and many other potential issues, as
reviewed extensively in the recent Intergovernmental Panel on Climate Change 4th Assessment
Report (IPCC 2007). Less discussed, but perhaps critical, are classic ideas in economic
development that link productivity to temperature (e.g., Montesquieu 1750; Marshall 1890;
Huntington 1915). Meanwhile, there are well-established, substantial effects of temperature on
mortality (e.g. Curriero et al. 2002; Deschenes and Moretti 2007; Deschenes and Greenstone
2007), temperature on crime (e.g. Field 1992; Jacob et al. 2007), and drought on conflict (Miguel
et al. 2004), all of which have direct and indirect effects on economic activity. Faced with these
different channels, the traditional approach to estimating the overall economic impact of climate
change is to use “Integrated Assessment Models” (IAM), which take some subset of
mechanisms, specify their effects, and then add them up (e.g. Mendelsohn et al. 2000, Nordhaus
and Boyer 2000, Tol 2002). Implementations of the IAM approach require many assumptions
about which effects to include, how each operates, and how they aggregate. 1
This paper takes a different approach. Rather than identifying mechanisms one-by-one
and summing up, we examine the effects of temperature and precipitation on a single aggregate
measure: economic growth. Specifically, we construct historical temperature and precipitation
data for each country and year in the world from 1950 to 2003 and combine this dataset with
historical growth data. The main identification strategy uses year-to-year fluctuations in
temperature and precipitation within countries to estimate the impact of temperature and
precipitation on economic growth. 2 This approach estimates the effect of short-run climate
fluctuations using relatively few assumptions. It examines aggregated outcomes directly, rather
than relying on a priori assumptions about what mechanisms to include and how they might
operate, interact, and aggregate.
Our main results show large, negative effects of higher temperatures on growth, but only
in poor countries. In poorer countries, we estimate that a 1◦C rise in temperature in a given year
reduces economic growth in that year by about 1.1 percentage points. In rich countries, changes
in temperature have no discernable effect on growth. Changes in precipitation also have no
substantial effects on growth in either poor or rich countries. We find broadly consistent results
across a wide range of alternative specifications.
As a result of these many assumptions, even primary users of these models acknowledge their substantial
limitations. For example, the Stern report (2007) describes IAM models as follows (p. 145):
“Making such estimates is a formidable task in many ways (discussed below). It is also a computationally
demanding exercise, with the result that such models must make drastic, often heroic, simplifications along
all stages of the climate-change chain. What is more, large uncertainties are associated with each element in
the cycle. Nevertheless, the IAMs remain the best tool available for estimating aggregate quantitative global
costs and risks of climate change.”
The use of annual variation in temperature and precipitation to estimate the impact of climate change was
pioneered by Deschenes and Greenstone (2007), who use annual county-level data in the United States to estimate
the impact of climate on U.S. agricultural output.
To interpret these effects, one can distinguish two potential ways temperature could
affect economic activity: 1) influencing the level of output, for example by affecting agricultural
yields, or 2) influencing an economy’s ability to grow, for example by affecting investments or
institutions that influence productivity growth. By looking at multiple lags of temperature, we
can examine whether temperature shocks appear to have temporary or persistent impacts on
economic output – and thus whether temperature has level or growth effects (or both). Our
results suggest that higher temperatures may reduce the growth rate in poor countries, not simply
the level of output. Since even small growth effects have large consequences over time, these
growth effects – if they persist in the medium run – would imply large impacts of temperature
We also find evidence for a broad set of mechanisms through which temperature might
affect growth in poor countries. While agricultural output contractions are part of the story, we
also find adverse effects of hot years on industrial output and aggregate investment. Further,
higher temperatures lead to political instability in poor countries, as evidenced by irregular
changes in the national leadership. These industry, investment, and institutional effects sit
outside the primarily agricultural focus of most economic research on climate change and
underscore the importance of an inclusive approach to understanding climate change
implications. These broader mechanisms also help explain how temperature might affect growth
rates in poor countries, not simply the level of output.
These results are identified using short-run fluctuations in temperature and precipitation.
A fundamental issue, however, is that the long-run effects of climate change may be quite
different from the effects of short-run fluctuations. For example, in the long run, adaptation
mechanisms might mitigate the short-run economic impacts that we observe. Alternatively,
climate change may have additional long-run effects, including changes in water tables, soil
quality, and sea level, producing larger impacts (IPCC, 2007; Meehl et al., 2004; Nicholls and
Although our approach (like others) cannot fully overcome these challenges, we can
make further headway by examining longer-term climate shifts. Mean global land temperatures
have risen nearly 1◦C since 1970 (Brohan et al. 2006), but countries have not warmed equally.
We therefore examine whether those countries that experienced the largest climate shifts
between early and late periods in our sample had the largest shifts in their growth rates. Though
this approach has less statistical power than using annual variation, the estimated effects of
increased temperature in poor countries over decade or longer time horizons are very similar to
our panel estimates. Should there be rapid adaptation to climate change, these estimates also
appear consistent with the overall cross-sectional relationship between temperature and percapita GDP found in the world today. To the extent that these historical effects continue, our
estimates suggest substantial negative consequences of climate change for poor countries and
few effects on rich countries.
The remainder of the paper is organized as follows. Section 2 introduces the data and
provides descriptive statistics. Section 3 describes the estimation strategy, presents the main
results, and considers a number of robustness checks. Section 4 considers channels that may link
climate change to national output. Section 5 estimates the effects of longer-run climate shifts,
offers projected implications of climate change using a standard climate model, and discusses the
limitations of such predictions. Section 6 concludes.
Data and Descriptive Statistics
The historical climate data is taken from the Terrestrial Air Temperature and
Precipitation: 1900-2006 Gridded Monthly Time Series, Version 1.01 (Matsuura and Willmott
2007). This data set provides worldwide (terrestrial) monthly mean temperature and precipitation
data at 0.5 x 0.5 degree resolution (approximately 56km x 56km at the equator). Values are
interpolated for each grid node from an average of 20 different weather stations, with corrections
We use geospatial software to aggregate the climate data to the country-year level. Our
main specifications use population-weighted average temperature and precipitation, where the
weights are constructed from 1990 population data at 30 arc second resolution (approximately
1km at the equator) from the Global Rural-Urban Mapping Project (Balk et al. 2004). We also
consider averaging based on geographic area, which produces broadly similar climate variables
for most countries. 3 Appendix I presents additional details about the climate data.
For economic data, we primarily use the Penn World Tables Version 6.2 (Heston et al.
2006). We also use data from the World Development Indicators (World Bank 2007) to examine
robustness and disaggregated value-added output from agriculture and industry. We focus on the
panel of 136 countries with at least 20 years of GDP data in the Penn World Tables, and consider
other samples as robustness checks.
Figure 1 presents population-weighted global mean temperature and precipitation from
1950 to 2006. The figure shows that the world has become about 1◦C warmer since the early
Countries where the weighting scheme makes a substantial difference are those with large, sparsely populated
areas with unusual climates: Russia (Siberia), Canada (the arctic and sub-arctic areas), the United States (Alaska),
and Australia (central Australia).
1970s, and that average precipitation has fallen by about 10 cm. The warming trend since the
1970s is well-documented (e.g. Brohan et al. 2006) and suggests a linear rate of change that,
should it continue, would predict an additional 3◦C warming by 2100, in line with many climate
models. The decline in precipitation is also well-documented, though this trend stands in contrast
to most climate models, which predict that global warming will come with increased
precipitation on average. 4
To examine variation in climate, Figure 2 summarizes temperature (left graph) and
precipitation (right graph) data for each country in the sample, plotted against log per-capita PPP
GDP in the year 2000. For each country, the circle symbols represent the mean levels of
temperature and precipitation in the first decade of our sample (1950-1959), the plus symbols
represent the mean levels in the last decade of our sample (1996-2005), and the gray lines
indicate the range of annual mean levels we observe for that country.
The left panel of Figure 2 shows the tremendous temperature variation across countries:
the hottest country in the world is Mauritania, with an average population-weighted temperature
of 28.4 ºC, and the coldest is Mongolia, with an average population-weighted temperature of
-1.77 ºC. Figure 2 also shows the strong relationship between temperature and per-capita income,
with hot countries tending to be poor and cold countries rich. This relationship has been known
since at least the 18th century (Montesquieu 1750) and has been further established using subnational data (Nordhaus 2006). The exceptions to this rule fall into two main groups: oil states of
the Middle East, such as Qatar and Kuwait, which are hot and wealthy, and Communist / postCommunist states, such as Mongolia and North Korea, which are cold and poor.
Historical area-weighted data for land shows a precipitation increase of nearly 1 cm over the 20th century.
However, a peak occurred in the 1950s, with precipitation falling across more recent decades (Neng et al., 2002;
New et al., 2001). On average, climate projections predict a rise in precipitation equal to about half of the recent
global decrease by the end of the 21st century (see IPCC 2007 Working Group 1 Chapter 10).
Looking at variability within countries, we see fluctuations in annual mean temperatures
on the order of about 2-3ºC. Thus, the max-min variation within countries is more than twice the
average increase in temperature observed over the period, and similar to the increase in global
temperatures expected to occur over the next century. Figure 2 further shows that, while there
tend to be larger temperature fluctuations in cooler countries, the upward trend in temperature
has occurred globally with similar magnitude in both hot and cold countries.
Examining the data on precipitation in the right panel of Figure 2 shows substantial
annual variability in precipitation in all but the very driest countries. However, there is no clear
relationship between the level of precipitation and the level of per-capita income in 2000.
To examine the variability further, Table 1 documents the extent of temperature and
precipitation fluctuations within countries. While the max-min difference in temperature is about
2-3ºC (Figure 2), a country’s temperature deviates more than 1ºC from the country mean
approximately once every fifteen years. Precipitation is more volatile, with deviations from mean
rainfall of about 400-500mm appearing once every fifteen years. When common global or
region-specific year fixed effects are removed, these deviations become somewhat more modest.
The effect of climate fluctuations on economic activity
In this section we develop the empirical framework for the analysis of climate shocks,
present our main results, and consider a variety of robustness checks.
Our empirical framework follows the derivation in Bond et al. (2007). To fix ideas,
consider the following simple economy: 5
Yit = e βTit Ait Lit
ΔAit / Ait = gi + γTit
where Y is aggregate output, L measures population, A measures labor productivity, and T
measures climate. Equation (1) captures the level effect of climate on production; e.g. the effect
of current temperature or precipitation on crop yields. Equation (2) captures the growth effect of
climate; e.g. the effect of climate on features such as institutions that influence productivity
Taking logs in the production function and differencing with respect to time, we have the
dynamic growth equation
git = g i + ( β + γ )Tit − βTit −1
where git is the growth rate of per-capita output. The “level effects” of climate shocks on output,
which come from equation (1), appear through β . The “growth effects” of climate shocks,
which come from equation (2), appear through γ . 6
We focus here on this simple production model. Appendix II extends the reasoning developed here to more general
dynamic panel models that incorporate richer lag structures and lagged dependent variables.
Rather than first-differencing (1), one could integrate (2), producing a fully-specified equation in the log level of
output. However, as Bond et al. (2007) notes, this creates non-stationarity in both output levels (on the left-hand
side) and accumulable factors (on the right-hand side). To avoid relying on cointegration assumptions for
identification, Bond et al recommend first-differencing.
The growth equation in (3) allows separate identification of level effects and growth
effects through the examination of transitory weather shocks. In particular, both effects influence
the growth rate in the initial period of the shock. The difference is that the level effect eventually
reverses itself as the climate returns to its prior state. For example, a temperature shock may
reduce agricultural yields, but once temperature returns to its average value, agricultural yields
bounce back. By contrast, the growth effect appears during the climate shock and is not reversed:
a failure to innovate in one period leaves the country permanently further behind. The growth
effect is identified in (3) as the summation of the climate effects over time.
The above reasoning extends to models where climate effects play out more slowly. 7
With more general lag structures in (1) and (2), the growth effect is still identified by summing
the lagged effects of the climate shock. This standard distributed-lag result is demonstrated
formally in Appendix II.
To estimate these effects, we run panel regressions of the form
g it = θ i + θ rt + ∑ j =0 ρ j Tit − j +ε it
where θ i are country fixed effects, θ rt are time fixed effects (interacted separately with region
dummies and a poor country dummy in our main specifications), ε it is an error term clustered by
country, and Tit is a vector of climate variables (temperature and precipitation) with up to L lags
included. In addition, we also consider variations of (4) that include interactions between climate
variables and country characteristics. We have verified using Monte Carlo analysis that the
For example, low temperatures in the latter part of one year could affect harv …
Purchase answer to see full
Our customer is the center of what we do and thus we offer 100% original essays..
By ordering our essays, you are guaranteed the best quality through our qualified experts.All your information and everything that you do on our website is kept completely confidential.
Academicwritingcompany.com always strives to give you the best of its services. As a custom essay writing service, we are 100% sure of our services. That is why we ensure that our guarantee of money-back stands, alwaysRead more
The paper that you order at academicwritingcompany.com is 100% original. We ensure that regardless of the position you are, be it with urgent deadlines or hard essays, we give you a paper that is free of plagiarism. We even check our orders with the most advanced anti-plagiarism software in the industry.Read more
The Academicwritingcompany.com thrives on excellence and thus we help ensure the Customer’s total satisfaction with the completed Order.To do so, we provide a Free Revision policy as a courtesy service. To receive free revision the Academic writing Company requires that the you provide the request within Fifteen (14) days since the completion date and within a period of thirty (30) days for dissertations and research papers.Read more
With Academicwritingcompan.com, your privacy is the most important aspect. First, the academic writing company will never resell your personal information, which include credit cards, to any third party. Not even your lecturer on institution will know that you bought an essay from our academic writing company.Read more
The academic writing company writers know that following essay instructions is the most important part of academic writing. The expert writers will, therefore, work extra hard to ensure that they cooperate with all the requirements without fail. We also count on you to help us provide a better academic paper.Read more