Rural Households in a Changing Climate


                  Javier E. Baez † Dorothy Kronick † Andrew D. Mason*


This paper argues that climate change poses two distinct, if related, sets of challenges for
poor rural households: challenges related to the increasing frequency and severity of
weather shocks and challenges related to long-term shifts in temperature, rainfall pat-
terns, water availability, and other environmental factors. Within this framework, we
examine evidence from existing empirical literature to compose an initial picture of
household-level strategies for adapting to climate change in rural settings. We find that
although households possess numerous strategies for managing climate shocks and
shifts, their adaptive capacity is insufficient for the task of maintaining—let alone im-
proving—household welfare. We describe the role of public policy in fortifying the ability
of rural households to adapt to a changing climate. JEL codes: Q12, Q54, O13



Introduction
How will rural households in developing countries adapt to climate change?
Rooted as it is in the hard sciences and in high-income communities, the public
dialogue on global environmental change has only recently begun to consider the
adaptive capacity of poor rural households. This inattention is not for lack of im-
portance; the welfare of these households concerns millions. Rather, it may be for
lack of definition: the issue is so broad and so complex that it seems to defy
focused consideration.
   One way to conceptualize adaptation in rural communities is to recognize two
effectively distinct phenomena associated with climate change: first, increased
weather variability and increased frequency and severity of extreme weather
events (broadly, “shocks”) and, second, gradual changes in temperature and rain-
fall patterns (broadly, “shifts”). From a climate perspective, these phenomena are
not actually distinct; from the perspective of adaptation among agricultural
households, however, dividing the issue into shocks and shifts provides a useful

The World Bank Research Observer
# The Author 2012. Published by Oxford University Press on behalf of the International Bank for Reconstruction and
Development / THE WORLD BANK. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
doi:10.1093/wbro/lks008           Advance Access publication November 14, 2012                               28:267–289
framework. Events that temporarily disrupt a household’s modal environmental
conditions can be considered shocks, and those that permanently change the
modal conditions can be considered shifts.
   Of course, the increasing frequency and severity of extreme weather events is
itself a shift because it permanently changes modal environmental conditions.
Likewise, gradual changes in temperature and rainfall patterns may manifest as
shocks (in the form of droughts and floods, for example). By categorizing the ways
in which rural households experience climate change, we clarify the relevance of
several established bodies of literature. This literature includes work on ambiguity
risk, agricultural technology adoption, reaction to weather-related information,
determinants of migration, income diversification, health, and past experiences
with permanent environmental changes. This paper synthesizes evidence from
this literature to develop a more inclusive and composite picture of household-
level strategies for adapting to climate change in rural settings.1 We also briefly
discuss the role of public policy in fortifying households’ capacity to adapt.
   We describe rural households’ numerous strategies for dealing with climate
change, and we outline the ways in which these strategies are insufficient for the
task of maintaining (let alone improving) household welfare. The threat to welfare
derives not only from the direct losses associated with climate shocks and shifts
but also from increased uncertainty, which complicates households’ attempts to
smooth consumption (with credit or insurance, for example), to smooth income
(by adjusting agricultural portfolios), and to evaluate options such as migration.
   The notion that some households may lack sufficient adaptive capacity or may
respond in ways that have adverse long-term consequences suggests a possible
role for public policy. To date, the dominant policy framework, to the extent that
one exists, has focused on what has been called a no-regrets approach to climate-
related adaptation policy (see Heltberg, Siegel, and Jorgensen 2009; de la Torre,
Fajnzylber, and Nash 2009; and World Bank 2010). In the face of the many un-
certainties associated with the future effects of climate change, the no-regrets ap-
proach emphasizes policies and investments today that can generate positive net
benefits “under all future scenarios of climate change and impacts” (Heltberg,
Siegel, and Jorgensen 2009). This approach has many merits, though there may
also be room for more specific action. Here, we draw attention to country experi-
ences with (i) better systems of information for adaptation and (ii) climate-
adapted risk management instruments (such as indexed production insurance),
and (iii) social safety nets that are targetable and scalable on the basis of weather
events.
   The remainder of this paper is organized as follows. Section II defines key con-
cepts related to rural households’ exposure, sensitivity, and capacity to adapt to
climate change. Drawing on extensive empirical literature on household responses
to weather shocks, section III analyzes how—and how well—households may be

268                                   The World Bank Research Observer, vol. 28, no. 2 (August 2013)
expected to cope with the increased frequency and severity of weather shocks as-
sociated with climate change. Drawing on several strands of empirical literature
on the determinants of household economic mobility, section IV explores how
households are likely to adapt to longer-term climatic shifts. Section V discusses
directions for public policy, and section VI concludes.




Exposure, Sensitivity, and Household Capacity
to Adapt to Climate Change
In the literature on climate change, the concept of vulnerability comprises three
elements: exposure (experience of climate conditions), sensitivity (response of the
physical environment to exposure, such as the response of crop yield to tempera-
ture and precipitation), and adaptive capacity (discussed below) (IPCC 2007 and
many others, notably Turner 2003 and Schro       ¨ ter, Polsky, and Patt 2005). This
section briefly describes rural households’ exposure and sensitivity to climate
change and then defines adaptive capacity, which is the subject of the remainder
of the paper.
   Substantial effort has been devoted to generating increasingly precise projec-
tions of climate trends and the exposure of rural households to these trends.
Summarizing this work is beyond the scope of this paper; below, we outline a few
points that are useful for our discussion of adaptive capacity. As noted above, we
organize our discussion according to two broad categories of climate changes:
shocks, the increase in weather variability and the frequency of extreme weather
events, and shifts, the long-term rise in global surface temperatures and related
gradual changes in sea level and rainfall patterns.
   Over the past century, global mean surface temperatures rose by 0.748C, on
average (the absolute global mean surface temperature is approximately 148C)
(IPCC 2007). The speed of warming accelerated as the century progressed. The
expected trajectory of warming over the next hundred years is a matter of consid-
erable debate. Existing models predict that global mean surface temperatures will
rise by an additional 1 to 58C by the year 2050.
   Global warming is one determinant of precipitation patterns and extreme
weather events, such as storms. Higher surface temperatures induce increased
evaporation from soils and bodies of water. The resulting elevated concentration
of water vapor leads to more precipitation in some places and more droughts in
others. Some precipitation that used to fall as snow now falls as rain. In the view
of many researchers, extreme weather events are more frequent and more severe
as a result of global warming; there are now more extreme warm nights, heat
waves, tropical storms, and hurricanes than there were 50 years ago. Figure 1

Baez et al.                                                                      269
illustrates a clear upward trend in the number of weather-related natural disas-
ters. The direction of the trend is clear, though the increase may be driven in part
by both better reporting practices and increasing human settlement in risk areas.
In regions such as Latin America and the Caribbean as well as East Asia and
the Pacific, the average annual frequency of disasters increased fivefold between
the 1970s and the 2000s and increased almost twofold in the most recent two
decades (during which changes in measurement techniques were less likely to
confound interpretation and inference).
    Agricultural production is often highly sensitive to the weather shocks and
shifts associated with climate change. Higher temperatures and changing rainfall
patterns may influence crop yields; in recent years, for example, climatic variabili-
ty related to El Nin˜ o Southern Oscillation (ENSO) affected wheat yields in Mexico,
cotton and mango growing cycles in Peru, and the incidence of plant diseases in
various other parts of Latin America. Cattle and dairy productivity may also
suffer: heat waves have been shown to negatively influence milk production in
Argentina. Water supply restrictions also reduce productive potential in agricul-
ture. Ocean acidification may damage fish stocks, and sea-level rise may eventual-
ly submerge coastal farmlands. Extreme weather events often destroy crops and
arable land.



Figure 1. Climate-related Disasters by Region (1970 –2009)




  Note: The count includes events that meet at least one of the following criteria: (1) 10 or more people reported
as dead, (2) 100 people reported as affected, (3) a declaration of a state of emergency, or (4) a call for
international assistance. LAC (Latin America and the Caribbean), SSA (Sub-Saharan Africa), ECA (East and
Central Europe), SAS (South Asia), EAP (East Asia and Pacific), MENA (Middle East and North Africa).
  Source: World Bank staff calculations based on EM-DAT: The OFDA/CRED International Disaster Database,
Catholic University of Louvain



270                                                The World Bank Research Observer, vol. 28, no. 2 (August 2013)
    This list is not exhaustive, of course. The point is merely that climate change
alters the natural conditions in which agricultural households operate. Although
employment in agriculture has fallen over the past half century as developing
economies have grown, hundreds of millions of people remain in the sector. For
nonagricultural households in rural areas, the environment is sensitive to climate
change: roads flood, infrastructure is damaged or destroyed, and diseases flourish
or fade.
    Perhaps because the term “adaptation” seems relatively straightforward, cli-
mate change authorities such as the Intergovernmental Panel on Climate Change
and the United Nations Framework Convention on Climate Change Secretariat
have not taken it upon themselves to sanction a common definition of the term in
the context of climate change (OECD 2006). However, some clarity of definition
may be useful for understanding what adaptation means in the context of our dis-
cussion. For the purposes of this paper, then, adaptation refers to household-level
actions taken (i) in response to observed or expected climate conditions and
(ii) with the objective of maintaining or improving household welfare.
    Adaptation, in this context, involves a multistage process among household de-
cision makers: first, signal detection, or the identification of a weather or climate
event; second, an evaluation of the expected consequences or impacts of that
event; third, a response, an observable change in household behavior that stems
from the evaluation; finally, feedback, or monitoring the outcomes of the adaptive
behavior (Kandlikar and Ribsey 2000). Planting alternative crops in response to
changing temperature, moving to a different location to avoid floods or droughts,
and insuring against weather-related asset losses could all be considered adaptive
actions within this process. Accordingly, we use the term adaptive capacity to
denote the extent to which households are able to undertake the processes and
actions required to maintain or improve welfare in the face of negative shocks and
shifts associated with climate change.
    Like the manifestations of climate change, the determinants of adaptive capaci-
ty among rural households are numerous. For the purposes of this article, these
determinants can be thought to fall into two broad categories: (i) economic
and physical factors and (ii) behavioral or psychological factors. The former cate-
gory includes factors traditionally discussed in development literature, such as
people’s levels of human capital, their access to credit, ownership and control of
assets (e.g., land, livestock), and physical mobility. The latter category involves in-
dividuals’ perceptions of risk and of their own adaptive capacity, which have been
the focus of more recent work (e.g., Bryan, Chowdhury, and Mobarak 2011). One
view of adaptation to climate change holds that these perceptions are just as im-
portant—perhaps more important—than physical and economic factors in deter-
mining adaptive capacity (Grothmann and Patt 2005).2



Baez et al.                                                                        271
Household Responses to Climate-change Shocks: Protecting
Consumption, Income, and Human Capital
Rural households have been exposed to weather shocks for millennia, and schol-
ars have studied households’ responses to these shocks for decades. In many
ways, household adaptation to weather shocks associated with climate change is
likely to mirror past adaptation to storms, floods, and droughts. In this sense, the
empirical evidence on household responses to past weather shocks is instructive.
At the same time, the increasing frequency and severity of weather shocks associ-
ated with climate change, if detected and evaluated, may produce qualitative
changes in household responses. Understanding these potential differences is key
to contextualizing existing evidence.
   First, the increasing frequency and severity of weather shocks will almost cer-
tainly alter the economic and physical factors that determine households’ adap-
tive capacity. Beyond direct damage to physical capital, more frequent and
extreme weather events may induce higher levels of precautionary savings,
which, in turn, may result in lower investment and a reduction in the long-term
income stream. Similarly, farmers may reduce overall productivity in an attempt
to mitigate risk by diversifying agricultural production portfolios in a suboptimal
manner. To our knowledge, there is little empirical evidence on the cumulative
effects of repeated weather shocks.
   We might consider the damage to physical assets and the necessity of addition-
al saving or risk-management activities to be quantitative or magnitude-based
changes in adaptive capacity. The key concept is that climate change will simply
scale up the adverse effects of weather shocks as we currently observe them.
   Perhaps more significant than these changes in scale, however, are the ways in
which increasing frequency and severity of weather shocks may imply qualitative
changes in the way that households respond to climate, principally by affecting
the behavioral and psychological determinants of adaptive capacity. In particular,
trends in climate shocks may reduce households’ ability to detect and appropriately
evaluate weather-related risk ex ante, thereby widening the gap between optimal
and actual adaptive behaviors. Recent theoretical and empirical work suggests
that ambiguity aversion may negatively affect investments in insurance, for
example (Bryan 2010). In Zimbabwe, farmers have been found to systematically
underestimate the risk of drought and to underestimate their own capacity to
adapt to it (Patt and Gwata 2002). This evidence is suggestive, although little is
known empirically about the way in which changes in the frequency or severity
of climate shocks affect the appropriateness of adaptive responses.
   Despite these potential differences between past and future adaptive behavior
and capacity, the existing literature on household responses to weather-related


272                                  The World Bank Research Observer, vol. 28, no. 2 (August 2013)
shocks can provide valuable insights into future adaptive behaviors. The central
message that emerges from previous work is that although households employ a
range of measures to manage risk and shocks, the (often informal) adaptation
mechanisms that they employ are generally insufficient for maintaining their
welfare in the face of weather shocks. Moreover, some coping mechanisms are
economically inefficient and may have adverse long-term consequences. As dis-
cussed above, we expect that the increasing frequency and severity of climate
shocks will only magnify these insufficiencies.


Consumption Smoothing
Weather shocks (as compared with household-specific shocks such as illness) pose
particular challenges for households intent on smoothing consumption. Because
they affect whole communities rather than individuals, they complicate local in-
surance or reciprocal gift-giving efforts, requiring risk pooling or intervention
across a broader set of geographic regions.
   To what extent will weather shocks associated with climate change reduce con-
sumption among rural households? A large body of evidence indicates that al-
though rural households’ efforts to smooth consumption in the face of weather
shocks can significantly mitigate negative effects, the success of these efforts
varies dramatically across both communities and households. In a survey of
papers based on longitudinal data from southern rural India, for example,
Morduch (2003) finds that the effect of transfers between households (i.e., gifts
given with the expectation of reciprocity, an active consumption-smoothing
strategy) ranges from a 40 percent reduction in total income risk to a 90 percent
reduction in total income risk.
   Similarly, a study on Thailand finds that variation in rainfall patterns pushed rural
households to save more in their efforts to smooth consumption (Paxson 1992).
However, a recent study of 162 villages in Ethiopia finds that although some house-
holds were able to protect consumption by selling livestock or borrowing from friends
and relatives, between 25 and 40 percent of respondents “did nothing” in response
to droughts, floods, and hailstorms (Deressa, Hassan, and Ringler 2009).
   Existing evidence also suggests that increased weather volatility associated with
climate change is likely to have the largest effects on the consumption of poor
rural households, which have the fewest assets, human resources, and social net-
works. A study from rural China finds, for example, that a 10 percent decline in
income is associated with a 4 percent decline in consumption among poor house-
holds but only a 1 percent decline among nonhouseholds (Jalan and Ravallion
1999). Likewise, evidence suggests that households with greater asset holdings
and members who are more educated are better able to protect themselves from
the impacts of shocks and to recover afterward. In rural Nicaragua and El

Baez et al.                                                                        273
Salvador, for example, recent studies show that households with higher levels of
education and greater asset holdings are both less likely to fall into poverty
during an aggregate shock and more successful at growing their incomes in the
aftermath (Vakis, Kruger, and Mason 2006; Beneke de Sanelı   ´u and Shi 2004).
   In sum, available evidence indicates that households’ abilities to smooth con-
sumption in the face of weather-related (and other) shocks are “real and signifi-
cant but not complete” (Murdoch 1995). The increasing weather volatility
associated with climate change is likely to further stretch the capacities of rural
agricultural households to protect their consumption and increase the risks of
their use of inefficient or costly mechanisms to respond to shocks. The evidence
also indicates that the increased frequency and severity of weather shocks is likely
to have the greatest impact among the poorest rural households.


Income Smoothing
In light of their inability to fully smooth consumption in the face of income
shocks, rural households dealing with decreasingly predictable returns to agricul-
ture may turn to income-smoothing strategies, such as altering crop portfolios or
diversifying sources of household income. In doing so, households often choose to
reduce risk even at the cost of significant foregone returns.
   One strategy adopted by households in the absence of formal mechanisms for in-
suring against production risk is to pursue low-risk, low-return agricultural portfo-
lios. In a rural municipality in southern Peru, for example, farmers plant and
cultivate many small, geographically dispersed and less profitable plots to reduce
aggregate production variance (Goland 1993). Likewise, farmers in southern India
whose production is vulnerable to rainfall variability are more likely to plant low-
risk, traditional varieties of rice and castor rather than higher-risk, high-yield
(more profitable) varieties (Morduch 1990). High levels of weather variability may
also discourage farmers from adopting production-enhancing technologies.
   There are significant costs associated with such forms of adaptation; conserva-
tive factor allocations may reduce profits and, thus, farmer incomes. In the
Peruvian municipality described above, the field-scattering strategy was found to
reduce farmers’ net yields by about 7 percent (Goland 1993). In the case of
southern rural India, microsimulation exercises showed that a one-standard devi-
ation reduction in weather risk (rainfall variability) would increase the average
profits for farmers at the bottom of the income distribution by nearly 35 percent
(Binswanger and Rosenzweig 1993).
   Another way that households adapt to increasingly unpredictable agroclimatic
conditions is to diversify their sources of income. Indian households that are vul-
nerable to transitory shocks, including those induced by weather risk, are more
likely than similar households to have a member employed in stable,

274                                   The World Bank Research Observer, vol. 28, no. 2 (August 2013)
nonagricultural sectors, even when off-farm employment offers lower wages
(Rosenzweig and Stark 1989). Studies from Guatemala, Thailand, and China,
among other places, describe similar behavior (see, e.g., World Bank 2005a;
Townsend 1995; Jalan and Ravallion 1999).
   Increasing weather variability and shocks could therefore lead to movements
away from agricultural labor markets. Although diversifying out of agriculture is
often associated with income growth and poverty reduction (World Bank 2007),
moving into nonagricultural employment in response to weather-related shocks is
not without its risks. A recent study using panel data from Bangladesh analyzes
the evolution of wages in the five years following a massive flood in 1998
(Mueller 2010). Although agricultural labor markets there experienced short-
term negative consequences due to the flood, nonagricultural labor markets per-
formed worse than agricultural labor markets in the medium term.


Human Capital Accumulation
Human capital accumulation is also highly susceptible to unfavorable weather
events. By reducing investment in education and health and by changing the inci-
dence of certain diseases, weather shocks may jeopardize long-term household
welfare, increasing susceptibility to future shocks and increasing the likelihood of
the intergenerational transmission of poverty.
   Households in rural India, for example, pull their children out of school to cope
with various weather-related income shocks (Jacoby and Skoufias 1997). In
Nicaragua, Hurricane Mitch negatively affected the progress of students in school
(Ureta 2005). There is some evidence that such transitory coping mechanisms
may solidify into permanent losses in human capital investment. Evidence from
Mexico shows that children who are withdrawn from school during difficult times
are one-third less likely to continue their studies than similar children whose
schooling is not interrupted (de Janvry, Finan, Sadoulet, and Vakis 2006). This
trend is particularly troubling in light of the fact that education—and female edu-
cation, in particular—appears to dramatically increase households’ adaptive ca-
pacity (see, for example, Blankespoor, Dasgupta, Laplante, and Wheeler 2010).
   These weather-related declines in schooling in rural areas are often associated
with increases in households’ use of child labor. In Tanzania, for example, there is
evidence that transitory income shocks created by accidental crop loss increased
child labor, more so among poor than nonpoor households (Beegle, Dehejia, and
Gatti 2003). In El Salvador, the fraction of children working outside the house-
hold following earthquakes in 2001 more than doubled, from 6.5 to 16.5 percent
(Santos 2007). Child labor also increased in rural Nicaraguan households affected
by the coffee crisis and in rural Nicaraguan households directly in the path of
Hurricane Mitch (Vakis, Kruger and Mason 2006; Baez and Santos 2007).

Baez et al.                                                                      275
   Weather-related shocks can also have a detrimental impact on the nutritional
status of poor rural women and children. Evidence from Ethiopia suggests that
households do not smooth consumption across individuals within the household;
rather, the nutritional status of women “bears the brunt of adverse shocks” (Dercon
and Krishnan 2000). Similarly, Baez and Santos (2007) find that Hurricane Mitch
increased the likelihood of early infant malnutrition by an alarming factor of four.
   As in the case of education, there is evidence that weather-induced shocks to
nutrition can have adverse long-term consequences on household welfare. In
Zimbabwe, for example, a recent study of young children who became stunted as
a result of a drought finds that those affected never fully recovered. Specifically,
the study found lasting negative effects not only on affected children’s body
weights but also on their school attainment and subsequent earnings (Alderman,
Hoddinott, and Kinsey 2006).
   Climate change may also affect health more directly by changing the incidence
of diseases. Malaria and dengue, for example, which are mosquito-borne diseases,
flourish in warm, humid conditions, depending on stagnant water (where mosqui-
toes breed) and wet air (in which adult mosquitoes can best survive). Cholera and
diarrheal diseases are also water-borne. Several studies have shown that the prev-
alence of these diseases increased in the wake of heavy rainfall caused by El Nin  ˜o
in various years in Ecuador, Bolivia, Peru, Colombia, and Venezuela (Vos, Velasco,
and Labastida 1999, WHO 1999, Bouma et al. 1997, Bouma and Dye 1997).
Extremes of heat and cold are also linked to higher morbidity and mortality,
mostly among the elderly (Rumel et al. 1993, Shumway, Azari, and Pawitan
1988, Bull and Morton 1978).




Household Adaptation to Long-term Shifts in Temperature
and Rainfall Patterns
Thus far, this paper has focused on how—and how well—rural households may
be expected to respond to shocks associated with climate change: tropical storms,
droughts, floods, and other extreme weather events. We now turn to gradual
shifts in temperature and rainfall patterns. Discrete weather-related shocks and
gradual shifts in temperature and precipitation levels associated with climate
change are, of course, intimately related. The effective distinction that we draw
here is that shocks represent deviations from a local mean, whereas shifts repre-
sent long-term changes in the mean.
  For rural households, these two phenomena require different types of adaptive
behaviors. Shocks require efforts to smooth consumption and income, whereas
shifts require longer-term—and perhaps permanent—changes in households’

276                                   The World Bank Research Observer, vol. 28, no. 2 (August 2013)
methods of income generation. Long-term climate shifts also complicate the task
of insuring against weather shocks because traditional insurance mechanisms are
designed to address variations in the mean of some variable of interest (e.g., agri-
cultural productivity or income) as opposed to secular changes in the means of
those variables over time.
   The existing evidence on farmers’ perceptions of climate change indicates that
farmers often recognize long-term shifts in environmental conditions and that
they attempt to respond to these shifts. In a recent survey conducted by IFPRI in
Sub-Saharan Africa, more than half of the respondents stated that they perceived
long-term increases in temperature and decreases in precipitation (Nhemachena
and Hassan 2007). They also indicated that a lack of credit and a lack of informa-
tion were significant barriers to adaptation.
   There are a number of ways in which households might try to adapt to
gradual, long-term changes in climactic conditions. Climate shifts may induce
households to (i) adapt their use of agricultural inputs and technologies; (ii) diver-
sify their sources of household income, either within agriculture or toward non-
agricultural enterprises; or (iii) emigrate from affected regions, all in an attempt
to compensate for the adverse effects of climate shifts on agricultural productivity,
farm profitability, and household income.
   Directly measuring how—and how effectively—households adapt to long-term
climate shifts is difficult because doing so requires long panels linking household
production behavior, productivity, and income to climate variables, data that
largely do not exist at present. Most panel evidence captures household economic
mobility and its determinants over relatively short periods of time, on the order of
five years.3 One recent study of a village in the Philippines uses a unique 30-year
panel of household data to examine the factors that enable household economic
mobility (Fuwa 2007). The study finds that in addition to economic growth
(which is found to be an important driver of household economic mobility), edu-
cation plays a critical role in upward economic mobility. Assets—specifically, land
holdings—are found to be important in preventing downward mobility among
rural farm households.4 The study does not examine the effect of changing
climate conditions, however.
   Despite the paucity of long-term panel data, a large body of empirical literature
on developing countries that examines adaptation to long-term economic changes
can inform our consideration of how well rural households—particularly poorer
households—will adapt to long-term climate shifts.


Adapting Agricultural Input Use and Adopting New Technologies
Adopting new technologies is one way in which rural agricultural households
can adapt to long-term climate shifts. A substantial body of empirical farm

Baez et al.                                                                       277
studies, some dating back several decades, examines determinants of farm produc-
tivity and profitability in developing countries across Asia, Africa, and Latin
America. The evidence indicates that not all households are equally capable of
adapting their use of productive inputs or adopting new technologies in the face
of changes in the economic environment. Studies consistently demonstrate the
importance of farmers’ education levels in facilitating higher farm productivity
through several channels: (i) direct improvements in worker productivity; (ii) in-
creased ability of workers and enterprise managers to make appropriate decisions
about the most efficient mix of inputs and technologies in production; (iii) in-
creased ability of decision makers to adapt to changing technologies, policies, and
the external economic environment; and (iv) at sufficiently high levels of educa-
tion, an ability to generate productivity-enhancing innovations.5


Diversifying Household Incomes
Diversifying household income sources, both within agriculture and toward non-
agricultural enterprises, is another form of adaptation that can potentially help
households manage longer-term climate shifts. The empirical literature suggests
that income diversification strategies are often at the heart of rural households’
ability to escape poverty (World Bank 2007). However, as the earlier discussion
on income smoothing suggests, not all households are equally capable of diversi-
fying their income sources in an efficient, welfare-enhancing way.
   Recent evidence using panel data from Nicaragua shows that family education
levels, wealth, control of physical and financial assets (specifically, land and
credit), more diversified income portfolios in the initial period, and greater access
and proximity to markets (and urban centers of nonagricultural activity) are all
positively associated with households’ economic mobility, income diversification,
and income growth (Vakis, Kruger, and Mason 2006). The findings from
Nicaragua are consistent with those from other recent panel studies. In
El Salvador, for example, households with higher levels of education, greater asset
holdings, and greater proximity to markets experienced faster real consumption
growth over the period analyzed (Beneke de Sanelı  ´u and Shi 2004).


Migration
A third form of adaptation that can potentially help households manage longer-
term climate shifts is migration away from affected regions, either to other rural
areas or to urban areas. Two studies on Brazil find that weather-related factors
currently induce emigration from affected areas (Assuncao and Chein Feres
2008, Mueller and Osgood 2009).

278                                   The World Bank Research Observer, vol. 28, no. 2 (August 2013)
   Consistent with the evidence on input use, technology adoption, and income
diversification, migration appears to be positively associated with people’s educa-
tion levels and income. In addition, the propensity to migrate appears to be posi-
tively associated with initial income levels. A recent study from Bangladesh
suggests that the poorest households may be least likely to migrate because
returns on migration are uncertain and because, relative to wealthier households,
the consequences of a negative outcome are severe. Paying for transport to the
city and then failing to obtain a job could mean starvation (Bryan, Chowdhury,
and Mobarak 2011).
   Similarly, analyses of Brazilian and Mexican census data indicate that illiterate
people are least likely to migrate, whereas those with higher education have a
higher probability of migrating than do other rural inhabitants (World Bank
2007). Higher schooling levels are also associated with higher probabilities of
migration from rural areas in China (Du, Park, and Wang 2005).



Directions for Public Policy
The fact that some rural households, particularly poorer ones, lack sufficient ca-
pacity to adapt to climate change suggests an important role for public policy. To
date, climate change studies that focus on household adaptation have tended to
take one of two approaches to strengthening household adaptation policies. The
first approach, common in agency reports, is to establish a comprehensive agenda
for action across sectors, often without attention to prioritization across sectors or
activities (see, e.g., IPCC 2007). The second, more recent discussion has focused
largely on what has been termed a no-regrets approach to fostering household ad-
aptation (see Heltberg, Siegel, and Jorgensen 2009; de la Torre, Fajnzylber and
Nash 2009; and World Bank 2009).
   The no-regrets approach emphasizes policies and investments that are part of a
sound development and poverty-reduction strategy for developing countries,
actions that can generate positive net benefits “under all future scenarios of
climate change and impacts” (Heltberg, Siegel, and Jorgensen 2009). Such an ap-
proach might include, for example, efforts to increase educational attainment
among members of rural households, improving poor households’ control of eco-
nomic assets (such as land and livestock), deepening of rural financial markets,
and improving rural households’ access to markets through further investments
in infrastructure.
   The evidence presented here suggests that even in the face of considerable uncer-
tainty regarding climate change—or perhaps precisely because of the uncertainties—
there may be occasion for more specific attempts to strengthen rural households’
adaptation capacity. We highlight three such areas here: (i) systems of information

Baez et al.                                                                       279
for adaptation, (ii) climate-adapted risk-management instruments (such as indexed
production insurance), and (iii) social safety nets that are targetable and scalable on
the basis of weather events.


Information for Adaptation
Because signal detection and the evaluation of the expected consequences are
critical to households’ ability to respond to climate change, developing systems of
information for adaptation will may improve households’ ability to adapt to both
weather shocks and long-term climate shifts.
   Indeed, better information on expected rainfall and seasonal temperatures,
coupled with appropriate agricultural extension advice, can help farm households
to make more appropriate decisions about crop mix, irrigation, the timing of
planting, and other ex ante production options. Since the major ENSO event of
1982/1983, various governments have attempted to affect the behavior of agri-
cultural households by producing and disseminating weather forecast information
(Magrin et al. 2007). In the northeastern Brazilian state of Ceara  ´ , for example, a
government agency engaged poor rural households in the Planting Time
Program. Households would receive high-yield seeds in exchange for a portion of
their crop; the agency would distribute the seeds at a time deemed most suitable
for planting, given seasonal weather forecasts. The agency also encouraged
households to plant drought-resistant crops before the low-rainfall season of
1991. Various assessments considered this intervention highly successful (Orlove
and Tosteson 1999). Likewise, in Mexico, ENSO forecasts have been used to adjust
crop mix (Magrin et al. 2007).
   The development of effective information systems for adaptation will face im-
portant implementation challenges. As Orlove and Tosteson (1999) note, the in-
stitutions that produce climate forecasts do not traditionally interact with the
institutional end users of those forecasts (such as food security planning agen-
cies), much less with households or individuals. Establishing relationships
between producers and users of climate forecasts can be complex. The Brazilian
agency mentioned above, for example, struggled with sharp public criticism after
issuing incorrect forecasts in the mid-1990s.
   In this context, Patt and Gwata (2002) identify six factors that are considered
critical if climate forecasts are to be effective in strengthening rural households’
ability to adapt to climate change. To induce the desired effect on farmer behavior,
climate information needs to be (1) considered credible, (2) communicated by
agents who are considered legitimate (that is, not beholden to special interests), (3)
calculated at appropriately detailed (that is, local) scales, (4) expressed in terms
that the audience can understand, (5) appropriately timed, and (6) relevant to the
decisions that farmers actively make. In this sense, the process of developing

280                                    The World Bank Research Observer, vol. 28, no. 2 (August 2013)
effective interventions to strengthen household adaptive capacity requires the
active participation of the households themselves (Schro¨ ter, Polsky, and Patt 2005).
   Although it is difficult to quantify the expected gains associated with develop-
ing information systems for adaptation, several studies have attempted to estimate
how past information dissemination programs have increased net returns to agri-
cultural production. Estimates range from 5 percent for Argentine soybeans
(Magrin and Travasso 2001) to 20 percent for Mexican maize (Jones 2001).
Attention to effective communication strategies and the effective use of new com-
munications technologies may serve to increase net returns associated with the
provision of information for adaptation. Efforts to disseminate weather and other
information relevant to production via cell phone text messages, a method that is
increasingly viable given the high and rising cell phone penetration rates in poor
rural areas, could help to expand outreach in some countries while lowering the
costs of disseminating information. Such approaches are already being used to
deliver market price information to farmers in developing countries.6


Climate-adapted Production Insurance
As discussed above, changes in average temperature and rainfall patterns over
time will require adaptation of the traditional insurance model to address the fact
that mean climate (and agronomic) conditions change in a systematic way over
time.7 More frequent and more intense weather shocks may also warrant a closer
link between climate variables and production insurance, perhaps building on a
small but a growing body of experience with weather-indexed insurance.
   Indeed, weather-indexed insurance may hold some potential for mitigating the
negative effects of climate change. In an attempt to address the moral hazard and
adverse selection issues that plague traditional crop insurance, some insurers now
offer products in which payouts are linked to the occurrence of a weather event,
such as low rainfall, which can be observed without a costly loss assessment pro-
cedure. Larson, Anderson, and Varangis (2004) provide a useful review of
weather-indexed insurance and related risk management tools. One difficulty,
they note, is that these products require reliable climate-related information and
therefore significant weather station infrastructure. Basis risk (that is, the risk
that index-based payouts will correspond only poorly with actual losses) may also
constitute a significant barrier to the purchase of index insurance products (e.g.,
Clarke 2011), although informal risk sharing that protects farmers against indi-
vidual-level idiosyncratic shocks may mitigate the problem of basis risk (Mobarak
and Rosenzweig 2012).
   Experience with weather-indexed insurance remains relatively limited, but
several recent applications are being monitored with interest to determine
whether they may have broader applicability in developing countries. For

Baez et al.                                                                       281
example, weather indices are currently being used for agricultural insurance in
Ontario and Alberta, Canada. In Argentina, in light of the positive correlation
between rainfall and milk yields, a rainfall insurance contract is being used by a
milk-producing cooperative. In Mexico, the agricultural reinsurance company
AGROSEMEX recently used weather derivatives to protect part of its crop reinsur-
ance portfolio exposed to weather risks.
   Index insurers cannot begin with a few clients and then expand (as in microfi-
nance). Because of the covariate nature of weather shocks, index insurers must
begin their operations with many clients in numerous areas (Morduch 2006). In
this context, the potential of weather-index insurance to strengthen adaptive ca-
pacity may be constrained by smallholder farmers’ demand for such instruments,
although recent work from India suggests that the availability of informal insur-
ance may increase the demand for index insurance under certain circumstances
(Mobarak and Rosenzweig 2012). Several pilot projects, in which information was
provided and smallholder farmers’ demand for insurance was “aggregated” by
local producer organizations and cooperatives, such as those in El Salvador and
Nicaragua, may provide lessons on how best to link such risk mechanisms to
those who need them (Varangis et al. 2003).


Social Safety Nets
Safety nets represent another important component of a risk management strat-
egy for climate change and are well suited to fortifying households’ ability to
adapt to both shocks and shifts. In the face of recent shocks (related both to
weather and economic shocks), it has become common to call for flexible safety
net programs that can respond to affected households’ needs without compromis-
ing governments’ long-term fiscal positions. Indeed, a number of developing coun-
tries around the world have adapted their safety net programs to help poor and
crisis-affected households deal with the recent food, fuel, and global financial
crises.
   One example of successful flexibility along these lines is the Honduras Social
Investment Fund (FHIS). Credited with limiting the rise of poverty in the wake of
Hurricane Mitch (which killed thousands of Hondurans, left one million homeless
and inflicted damage equivalent to two-thirds of GDP), the FHIS was a public
program created in 1990 to finance small-scale investments in poor communities.
Originally conceived as a way to address the social impacts of structural adjust-
ment policies, the FHIS nimbly transformed into an emergency response program
after Mitch devastated the country in 1998. Within 100 days of the hurricane,
the program approved $40 million for 2,100 community projects; by the end of
1999, the FHIS had financed 3,400 projects, four times the number financed in a
comparable prehurricane period (Warren 2003).8

282                                  The World Bank Research Observer, vol. 28, no. 2 (August 2013)
   There have been more recent experiences with safety net programs helping to
protect the poor from shocks, including weather shocks. In 2000, Nicaragua es-
                                                                    ´ n Social (RPS).
tablished a conditional cash transfer program called Red de Proteccio
Although this program was developed to address structural poverty and to foster
human capital development among the poor, the RPS was also effective in reduc-
ing the vulnerability of poor rural households to income shocks. During a sharp
downturn in world coffee prices from 2000 to 2003, consumption among coffee-
growing households that were RPS beneficiaries declined by only 2 percent,
whereas consumption among coffee-growing households who were not beneficia-
ries declined by 30 percent (Vakis, Kruger, and Mason 2006). Building on the
success of the RPS, the government of Nicaragua introduced a cash-transfer pilot
program in a drought-prone region. After a mudslide in eastern Nicaragua, the
program quickly responded by transferring resources to all affected households.9
An evaluation of the pilot program found that the income and consumption of
beneficiaries was significantly more resilient to droughts, price shocks, and health
shocks than that of nonbeneficiary families.
   One key feature of the Nicaragua pilot was targeting the program, ex ante, to
geographic regions of the country that faced elevated levels of weather-related risk
(specifically drought risk, in this case).10 Indeed, as with climate-adapted insur-
ance mechanisms, safety net programs designed to protect households from in-
creasingly frequent and severe weather shocks must address the increased
information-related challenges associated with targeting on the basis of weather
risk. Recent advances in mapping technology, combined with increasingly precise
weather-risk forecasts, can help policy makers to address these issues.



Conclusion
Under the best of circumstances, agricultural livelihoods are difficult to sustain
and improve. Even in the friendliest of agronomic environments, small-scale
farming can be incredibly challenging and risky. For those rural households in de-
veloping countries that depend on agriculture, climate change adds yet another
dimension of difficulty to the problem of maintaining and improving welfare.
The same is true, albeit via different mechanisms, for rural households whose
income does not come primarily from agricultural activity.
   This paper has argued that climate change poses two distinct, if interrelated,
sets of challenges for poor rural households: challenges related to the increasing
frequency and severity of weather shocks and those related to long-term shifts in
temperature, rainfall patterns, water availability, and other environmental factors.
Although much is unknown about how poor rural households will adapt to these
shocks and shifts, a substantial body of empirical evidence from developing

Baez et al.                                                                       283
countries provides a reasonably robust sense of the strengths and limitations of
available adaptation mechanisms.
   An important and encouraging finding from the literature is that rural
households—even poor ones—possess a number of strategies for mitigating the
effects of shocks associated with climate change. They may attempt to smooth
consumption in the face of income shocks (using credit, insurance, savings, and
other methods), and they may attempt to smooth income (by adjusting agricultural
portfolios or diversifying sources of income, for example). The evidence indicates,
however, that households’ abilities to smooth consumption and income are less
than perfect. Moreover, some strategies, such as employing child labor, come with
clear risks.
   Longer-term climate shifts create other challenges to rural households’ liveli-
hoods and call for other types of adaptation responses. Rural agricultural house-
holds confronted with significant environmental and agronomic changes due to
climate change will need to adjust their livelihood strategies if they are to main-
tain or increase their incomes. Adapting to climate shifts may require farming
households to adjust their cropping choices and alter their use of productive
inputs and technologies. Farmer households may also seek to diversify their
income sources or to migrate to areas less affected by long-term changes in tem-
peratures and precipitation.
   Understanding the factors that affect rural households’ capacity to adapt to
shocks and shifts will be important in designing effective interventions to
strengthen their adaptation capacity. Here, the evidence provides some consistent
messages: policies and programs that help to increase households’ human capital
and improve their access to physical assets, financial capital, and markets can
play important roles in strengthening their capacity to adapt to climate change.
In this sense, many interventions that can fortify rural households’ adaptive ca-
pacity are also part of an effective long-term development and poverty-reduction
strategy for developing countries. Hence, there is an appeal to a no-regrets ap-
proach that focuses on interventions with positive net benefits regardless of how
the shocks and shifts associated with climate change unfold.
   Yet, perhaps precisely because of the inherent uncertainties of climate change,
an intense focus on the adaptive capacity of rural households warrants consider-
ation. We have discussed three potential focus areas. In addition, investing in im-
proved household data collection and analysis, including the development of
longer-term panels of data that can facilitate direct observation and analysis of
households’ and policy makers’ responses to climate-related shocks and shifts,
will ultimately improve our understanding of adaptation to climate change in
rural areas.




284                                  The World Bank Research Observer, vol. 28, no. 2 (August 2013)
Notes
*Andrew D. Mason is with the Poverty Reduction and Economic Management Department, East
Asia and Pacific Region, The World Bank, Washington, DC 20433. Javier Baez is with the Poverty
Reduction and Economic Management Department, Latin America and Caribbean Region, The
World Bank, Washington DC 20433. Email: jbaez@worldbank.org. Dorothy Kronick is at Stanford
University. Email: dkronick@stanford.edu. An earlier version of this article was written as a back-
ground paper to the World Bank study for Latin America and the Caribbean, Low Carbon, High
Growth: Latin American Responses to Climate Change. The authors gratefully acknowledge support and
feedback from Augusto de la Torre, Pablo Fajnzylber, and John Nash, the lead authors of that
report, as well as Helena Ribe from the Bank’s Human Development Department for Latin America
and the Caribbean. We also thank three anonymous referees for their comments. The corresponding
author is Andrew D. Mason; email address: amason@worldbank.org. The findings, interpretations,
and conclusions expressed in this paper are entirely those of the authors. They do not necessarily
represent the views of the International Bank for Reconstruction and Development/World Bank and
its affiliated organizations, the Executive Directors of the World Bank, or the governments that they
represent.
    1. Although climate change creates benefits and opportunities for rural households in regions
where expected shifts may actually improve local agronomic conditions, we focus here on household
adaptation to expected adverse shocks and shifts.
    2. In the context of climate change, behavioral factors will likely be linked to economic factors
through human capital in the sense that people’s perceptions of risk and, even, of their own adap-
tive capacity are a function of their ability to detect and evaluate the actual (or likely) impacts of a
weather/climatic event. Although observing and evaluating the impact of shocks may be relatively
straightforward, assessing the risks associated with shifts can be much more complex.
    3. Overall, panel studies support the notion that education, assets, access to finance, and access
to markets (to different degrees in different places) are important for household income growth and
the movement of the poor out of poverty (see Dercon and Shapiro 2007 for a review of evidence).
Dercon and Shapiro (2007) also note that risk and shocks serve to inhibit household income
growth and mobility. In that sense, households’ vulnerability to shocks can affect their ability to
adapt to long-term climate shifts.
    4. Interestingly, among smallholder farmers, larger landholdings are found to inhibit movement
toward nonagricultural employment, perhaps by strengthening these farmers’ ties or commitment
to farming activities.
    5. Among the studies presenting empirical findings on the relationship between farmer educa-
tion and agricultural productivity, including the effect of education on input use and technology
adoption, see Lockheed et al. 1980; Jamison and Lau 1982; Moock and Addou 1994; Hussain and
Byerlee 1995; and World Bank 2007.
    6. For example, seminomadic herders in rural Mongolia can now subscribe to a service through
which they can obtain market price data on key commodities in the nearest district towns through
their cell phones.
    7. It should be noted that this, by itself, argues for better, more accessible, information on local
shifts in temperature and rainfall levels and may generate incentives for the production and dissemi-
nation of such information, independent of the perceived demand among rural households.
    8. Projects prioritized clearing debris and repairing or rebuilding water lines, sanitation systems,
roads, bridges, health centers, and schools, thus hastening national recovery and generating approx-
imately 100,000 person-months of employment in the three months following the crisis. The decen-
tralized structure and institutional flexibility of the FHIS facilitated its brisk and influential response.
Building on strong preexisting partnerships with municipalities and communities, FHIS directors es-
tablished 11 temporary regional offices and quickly delegated resources and responsibilities.



Baez et al.                                                                                           285
Directors reduced the number of steps in the subproject cycle from 50 to 8, established safeguards to
ensure accountability and transparency, and effectively accessed International Development
Association financing.
   9. The transfers took place within three months of the event; by the sixth month after the event,
RPS had enrolled all affected households as beneficiaries.
  10. Beneficiaries were selected from among RPS enrollees using a proxy-means test.




References
Alderman, H., J. Hoddinott, and B. Kinsey. 2006. “Long Term Consequences of Early Childhood
   Malnutrition.” Oxford Economic Papers 58 (3): 450 –74.
Assuncao, J., and F. Chein Feres. 2008. “Climate Migration.” Background Paper for World Bank
   Flagship Report on Climate Change in Latin America, Low Carbon, High Growth: Latin American
   Responses to Climate Change, World Bank Latin American and Caribbean Studies.
Baez, J., and I. Santos. 2007. “Children’s Vulnerability to Weather Shocks: A Natural Disaster as a
   Natural Experiment.” Ph.D. diss. Syracuse University, Maxwell School of Public Affairs, Syracuse,
   NY, and Harvard University, Kennedy School of Government, Cambridge, MA.
Beegle, K., R. Dehejia, and R. Gatti. 2003. “Child Labor, Crops Shocks and Credit Constraints.”
   NBER Working Paper 10088. National Bureau of Economic Research, Cambridge, MA.
Beneke de Sanfeliu, M., and M. Shi. 2004. Dina  ´ mica Del Ingreso Rural En El Salvador. 1st ed. San
                                 ´ n Salvadoren
   Salvador, El Salvador: Fundacio            ˜ a para el Desarrollo Econo´ mico y Social, FUSADES,
   2004.
Binswanger, H., and M. Rozensweig. 1993. “Wealth, Weather Risk and the Composition and
   Profitability of Agricultural Investments.” The Economic Journal 103:56 –78.
Blankespoor, B., S. Dasgupta, B. Laplante, and D. Wheeler. 2010. “The Economics of Adaptation to
   Extreme Weather Events in Developing Countries.” Working Paper 199. Center for Global
   Development, Washington, DC.
                                                                     ˜ o in Venezuela.” Journal of
Bouma, M., and C. Dye. 1997. “Cycles of Malaria Associated with El Nin
  the American Medical Association 278:1772 –74.
Bouma, M., G. Poveda, W   . Rojas, D. Chavasse, M. Quinones, J. Cox, and J. Patz. 1997. “Predicting
  High-Risk Years for Malaria in Colombia Using Parameters of El Nin     ˜ o Southern Oscillation.”
  Tropical Medicine and International Health 2:1122 –27.
Bryan, G. 2010. “Ambiguity and Insurance.”Ph.D. diss. Yale University, Department of Economics,
   New Haven, CT. (http://www.asb.unsw.edu.au/schools/economics/Documents/G.%20Bryan%20-%
   20Ambiguity%20and%20Insurance.pdf).
Bryan, G., S. Chowdhury, and A. M. Mobarak. “Seasonal Migration and Risk Aversion.” Discussion
   Paper DP8739. Yale University, Department of Economics, New Haven, CT. (http://faculty.som.yale.
   edu/mushfiqmobarak/migration.pdf).
Bull, G.M., and J. Morton. 1978. “Environment, Temperature, and Death Rates.” Age and Aging
   7 (4): 210 –24.
Clarke, Daniel. 2011. “A Theory of Rational Demand for Index Insurance.” Discussion Paper Series
   572. Department of Economics, Oxford University, Oxford.
De Janvry, A., F. Finan, E. Sadoulet, and R. Vakis. 2006. “Can Conditional Cash Transfers Serve as
   Safety Nets in Keeping Children at School and from Working when Exposed to Shocks?” Journal
   of Development Economics 79 (2): 349–73.


286                                          The World Bank Research Observer, vol. 28, no. 2 (August 2013)
De la Torre, A., P. Fajnzylber, and J. Nash. 2009. Low Carbon, High Growth: Latin American Responses
   to Climate Change. World Bank, Washington DC.
Dercon, S., and P. Krishnan. 2000. “In Sickness and in Health: Risk-Sharing Within Households in
   Rural Ethiopia.” Journal of Political Economy 108 (41): 688–727.
Dercon, S., and J.S. Shapiro. 2007. “Moving on, Staying Behind, Getting Lost: Lessons on Poverty
   Mobility from Longitudinal Data.” In D. Narayan, and P      . Petesch, eds., Moving Out of Poverty:
   Cross-Disciplinary Perspectives on Mobility, pp. 77 –126. Washington DC and New York: A co-pub-
   lication of the World Bank and Palgrave Macmillan.
Deressa, T., R. Hassan, and C. Ringler. 2009. “Household Vulnerability to Climate Change: The Case
   of Farmers in the Nile Basin of Ethiopia.” Discussion Paper 00935. International Food Policy
   Research Institute, Washington, DC.
Du, Y., A. Park, and S. Wang. 2005. “Migration and Rural Poverty in China.” Journal of Comparative
  Economics 33:688 –709.
Fuwa, N. 2007. “Pathways Out of Rural Poverty: A Case Study in Socio-economic Mobility in the
  Rural Philippines.” Cambridge Journal of Economics 31:123 –44.
Goland, C. 1993, “Field Scattering as Agricultural Risk Management: A Case Study from Cuyo
   Cuyo, Department of Puno, Peru.” Mountain Research and Development 13 (4): 317 –38.
Grothmann, T., and A. Patt. 2005. “Adaptive capacity and human cognition: The process of individ-
   ual adaptation to climate change.” Global Environmental Change 15:199 –213.
Heltberg, R., P. Siegel, and S. Jorgensen. 2009. “Addressing Human Vulnerability to Climate Change:
   Toward a ‘No Regrets’ Approach.” Global Environmental Change 19:89– 99.
Hussain, S.S., and D. Byerlee. 1995. “Education and Farm Productivity in Post-Green Revolution
  Asia.” In G.H. Peters, and D. Headley, eds., Agricultural Competitiveness, Market Forces, and Policy
  Choices. Proceedings of the 22nd Conference of the International Agricultural Economics
  Association, Aldershot, UK: Dartmouth Publishers.
Intergovernmental Panel on Climate Change (IPCC). 2007. Climate Change 2007: Impacts, Adaptation
   and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the
   Intergovernmental Panel on Climate Change. eds. M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der
   Linden, and C.E. Hanson. Cambridge, UK: Cambridge University Press.
Jacoby, H., and E. Skoufias. 1997. “Risk, Financial Markets, and Human Capital in a Developing
   Country.” Review of Economic Studies 64 (3): 311 –35.
Jalan, J., and M. Ravallion. 1999. “Are the Poor Less Well Insured? Evidence on Vulnerability to
    Income Risk in Rural China.” Journal of Development Economics 58 (1): 61 –81.
Jamison, D. T., and L. Lau. 1982. Farmer Education and Farm Efficiency. Baltimore: Johns Hopkins
   University Press.
Jones, J. 2001. “Comparative Assessment of Agricultural Uses of ENSO-based Climate Forecasts in
   Argentina, Costa Rica and Mexico: Project Summary.” IAI Initial Science Program – III Project,
   University of Florida, Gainsville, FL.
Kandlikar, M., and J. Ribsey. 2000. “Agricultural Impacts of Climate Change: If Adaptation is the
  Answer, What is the Question?” Climate Change 45:529– 39.
Kochar, A. 1999. “Smoothing Consumption by Smoothing Income: Hours-of-Work Responses to
   Idiosyncratic Agricultural Shocks in Rural India.” The Review of Economics and Statistics 81 (1): 50–61.
Larson, D., J. Anderson, and P  . Varangis. 2004. “Policies on Managing Risk in Agricultural
   Markets.” The World Bank Research Observer 19 (2): 199–230.
Lockheed, M.E., T. Jamison, and L. J. Lau. 1980. “Farmer Education and Farm Efficiency: A Survey.”
   Economic Development and Cultural Change 29 (1): 37– 76.


Baez et al.                                                                                            287
Magrin, G., C. Gay Garcia, D. Cruz Choque, J. C. Gimenez, A. R. Moreno, G. J. Nagy, C. Nobre,
  and A. Villamizar. 2007. “Latin America.” In M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der
  LindenC.E. Hanson, eds., Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution
  of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate
  Change, pp. 582–615. Cambridge, UK: Cambridge University Press.
Magrin, G., and M. Travasso. 2001. “Economic Value of ENSO-based Climatic Forecasts in the
  Agricultural Sector of Argentina.” 2nd International Symposium Modelling Cropping Systems,
  European Society of Agronomy (ESA), Florence, Italy: 139 –40.
Mobarak, A. Mushfiq, and Mark Rosenzweig. 2012. “Selling Formal Insurance to the Informally
  Insured.” Working Paper 97. Yale University, Department of Economics, New Haven, CT. (http://
  faculty.som.yale.edu/mushfiqmobarak/insurance.pdf ).
Moock, P.R., and H. Addou. 1994. “      Agricultural Productivity and Education.” In T. Husen, and
  N. Postlethwaite, eds., International Encyclopedia of Education Vol. 1. Oxford: Pergamon Press.
Morduch, J. 1995. “Income Smoothing and Consumption Smoothing.” Journal of Economic Literature
  37 (4): 1569–1614. In S. Dercon, ed., Insurance Against Poverty. Oxford: Oxford University Press.
        2006. “Micro-insurance: The next revolution?” In A. Banerjeee, R. Be          ´ nabou, and
   D. Mookherjee, eds., Understanding Poverty, pp. 337–356. Oxford: Oxford University Press.
Mueller, V. 2010. “Short- and Long-Term Effects of the 1998 Bangladesh Flood on Rural Wages.”
  Discussion Paper 00956. International Food Policy Research Institute, Washington, DC.
Mueller, V., and D. Osgood. 2009. “Long-Term Consequences of Short-term Precipitation Shocks:
  Evidence from Brazilian Migrant Households.” Agricultural Economics 40:573 –586.
National Institute of Environmental Health Sciences. 1999. “Environmental Impacts of Hurricane
  Mitch.” Environmental Health Perspectives 107 (3): A139 –40.
Nhemachena, C., and R. Hassan. 2007. “Micro-Level Analysis of Farmers’ Adaptation to Climate
  Change in Southern Africa.” Discussion Paper 00714. International Food Policy Research
  Institute, Washington, DC.
Organisation for Economic Cooperation and Development (OECD; E. Levina and D. Tirpak). 2006.
   “Key Adaptation Concepts and Terms.” Draft paper, Agenda Document 1, OECD/IEA Project for
   the Annex 1 Expert Group on the UNFCCC.
Orlove, B., and J. Tosteson. 1999. “The Application of Seasonal to Interannual Climate Forecasts
   Based on El Nin ˜ o-Southern Oscillation (ENSO) Events: Lessons from Australia, Brazil, Ethiopia,
   Peru, and Zimbabwe.” Working Paper 99-3. Berkeley Workshop on Environmental Politics,
   University of California, Berkeley.
Patt, A., and C. Gwata. 2002. “Effective seasonal climate forecast applications: Examining con-
   straints for subsistence farmers in Zimbabwe.” Global Environmental Change 12:185 –95.
Paxson, C. 1992. “Using Weather Variability to Estimate the Response of Savings to Transitory
   Income in Thailand.” American Economic Review 82 (1): 15 –34.
Rosenzweig, M., and O. Stark. 1989. “Consumption Smoothing, Migration, and Marriage: Evidence
   from Rural India.” Journal of Political Economy 97 (4): 905–26.
Rumel, D., L. Riedel, M. Latorre, and B. Duncan. 1993. “Infarto do Miocardo e Acidente Vascular
  Cerebral Associados a Alta Temperatura e Monoxido de Carbono em Area Metropolitana do
  Sudeste do Brasil.” Revista de Saude Publica 27 (1): 15 –22.
Santos, I. 2007. “Risk-Sharing and the Role of Inter-Household Transfers after a Natural Disaster.
   Evidence from Hurricane Mitch in Nicaragua.” Ph.D. diss. Kennedy School of Government,
   Harvard University, Cambridge, MA.
    ¨ ter, D., C. Polsky, and A. Patt. 2005. “Assessing Vulnerabilities to the Effects of Global Change:
Schro
   An Eight-Step Approach.” Mitigation and Adaptation Strategies for Global Change 10:573 –96.


288                                            The World Bank Research Observer, vol. 28, no. 2 (August 2013)
Shumway, R., A. Azari, and Y. Pawitan. 1988. “Modeling Mortality Fluctuations in Los Angeles as
  Functions of Pollution and Weather Effects.” Environmental Research 45:224– 41.
Timmins, C. 2003. “Health and Climate Change in the Developing World: The Impact of Global
   Warming on Brazilian Morbidity Patterns.” Working Paper. Duke University, Durham, NC.
Townsend, R. 1994. “Risk and Insurance in Village India.” Econometrica 62 (3): 539– 91.
         1995. “Financial Systems in Northern Thai Villages.” The Quarterly Journal of Economics 110
    (4): 1011–46.
Turner, B.L., R.E. Kasperson, P.A. Matson, J.J. McCarthy, R.W. Corell, L. Christensen, N. Eckley, J.X.
  Kasperson, A. Luers, M.L. Martello, C. Polsky, A. Pulsipher, and A. Schiller. 2003. “A Framework
  for Vulnerability Analysis in Sustainability Science.” Proceedings of the National Academy of
  Sciences, United States of America 100:8074 –9.
Ureta, M. 2005. “Hurricane Mitch, Family Budgets and Schooling in Nicaragua.” Working Paper.
   Department of Economics, Texas A&M University, College Station, Texas.
Vakis, R., D. Kruger, and A. Mason. 2006. “Shocks and Coffee: Lessons from Nicaragua.” In Income
   Stabilization in Agriculture: The Role of Public Policies, Proceedings of the 86th EAAE Seminar,
   Anacapri, Italy: 171– 231.
Varangis, P., P. Siegel, D. Giovannucci, and B. Lewin. 2003. “Dealing with the Coffee Crisis in
   Central America: Impacts and Strategies.” World Bank Policy Research Working Paper 2993.
   World Bank, Washington, DC.
Vos, R., M. Velasco, and E. Labastida. 1999. “Economic and Social Effects of El Nino in Ecuador,
   997-1998.” Inter-American Development Bank, Sustainable Development Department, Technical
   Paper Series, Washington, DC.
Warren, D. 2003. “Coping with a Natural Disaster: Hurricane Mitch and the Honduran Social
  Investment Fund.” Spectrum Magazine, Fall: 20 –21.
World Bank. 2005a. “Shocks and Social Protection: Lessons from the Central American Coffee
  Crisis.” Central America Department, Latin America and Caribbean Region, Washington, DC.
       . 2005b. “DR-CAFTA: Challenges and Opportunities for Central America.” Central America
    Department and Office of the Chief Economist, Latin America and Caribbean Region,
    Washington, DC.
       . 2007. “World Development Report 2008: Agriculture for Development.” World Bank,
    Washington, DC.
       . 2010. “World Development Report 2010: Development and Climate Change.” World Bank,
    Washington, DC.
World Health Organization (WHO). 1999. “El Nin  ˜ o and Health – Protection of the Human
  Environment.” Protection of the Human Environment: Task Force on Climate and Health, WHO/
  SPE/HE/99.4, Geneva.




Baez et al.                                                                                       289