The Heat Is Online

The Climate Crisis and the Adaptation Myth -- Robert Repetto

The Climate Crisis and the Adaptation Myth

Robert Repetto

Working Paper Number  13, Fall 2008


The Climate Crisis and the Adaptation Myth

Robert Repetto

Professor in the Practice of Economics (retired)

Yale School of Forestry & Environmental Studies


i. the issues


Influential studies have predicted that moderate climate change, up to 3 or 4 degrees Fahrenheit, will not be very damaging to the United States as a whole and will bring some benefits. Underlying the argument that climate change will not be very damaging to the U.S. economy is the contention that vulnerable organizations, firms and households will take steps to adapt. This assumption is based partly on the fact that the United States is rich in technology, economic resources, competent organizations and educated people, all of which combine to create a high capacity to adapt. More fundamentally, the contention rests on the observation that the United States spans a wide variety of climatic conditions to which households and enterprises have adapted successfully in the past. According to a recent review, The literature indicates that U.S. society can on the whole adapt with either net gains or some costs if warming occurs at the lower end of the projected range of magnitude, assuming no change in climate variability and generally making optimistic assumptions about adaptation.


These are key assumptions. The perception that damages will be limited has been a significant factor in the conclusion reached by the U.S. government that it was not in the national interest to join in the Kyoto Protocol agreement to reduce carbon emissions because the benefits in damages averted would be small relative to the costs incurred in mitigating GHG emissions in the U.S. to the extent called for in the Protocol. Furthermore, refusal by the United States, the worlds richest country and largest cumulative emitter of greenhouse gases, to ratify the Kyoto Protocol discouraged other countries from joining or implementing an international agreement to limit emissions. Consequently, assumptions regarding adaptation in the U.S. have had broad policy repercussions.


Are these assumptions justified? Certainly the potential for adaptation exists. Studies have identified steps in key vulnerable sectors that can significantly reduce damages. The National Assessment identifies specific steps in all regions and sectors that could prevent or limit damages. Such studies have indicated that damages to agriculture, forestry and other economic activities can be greatly reduced if economic agents adapt efficiently. For example, damages to agriculture are estimated to be 50 percent less as a result of farmer adaptation. Economists have criticized the dumb farmer  assumption that farmers will just suffer damages and not do anything about it. Other studies have estimated that damages from sea level rise could be reduced by over half, even including the costs of adaptation, if appropriate protective measures are taken.  


However, saying that the U.S. can adapt does not imply that it will adapt, at least not in the efficient and timely way needed if major damages are to be avoided. The question is whether it is likely that such steps will actually be taken and whether they will be taken in sufficient time to limit damages. If not, damages from climate change will be considerably higher than has been estimated. There is an important distinction between anticipatory or preventive adaptation that predicts and responds to vulnerabilities before damages are incurred and reactive adaptation that gears up to limit the recurrence of damage only after effects of climate change have been felt and damage done, in order to limit recurrence of the damage.


If adaptation is mainly reactive, then damages will be much greater. Unfortunately, experience shows that, in the United States, responses to disaster are mainly reactive, often characterized by inattention beforehand and over-response afterwards. In the case of climate change, reactive adaptation will be especially costly because, decade by decade, the severity of climate change impacts is likely to increase as greenhouse gas concentrations in the atmosphere rise. Reactive adaptation would be likely to lag persistently behind the emerging risks. The more rapid the rise in atmospheric concentrations, the faster the rate of climate change and the less effective reactive adaptation is likely to be.


Reactive adaptation would be likely to lag persistently behind the emerging risks. The more rapid the rise in atmospheric concentrations, the faster the rate of climate change and the less effective reactive adaptation is likely to be.


Moreover, it is possible that whatever pro-active adaptive measures are put in place may be partially or wholly offset by maladaptations, which serve to increase vulnerability and the likelihood of future damages. One example is continuing shoreline and floodplain real estate development. The International Hurricane Research Center has identified coastal regions in the United States that are most vulnerable to future hurricane damage. Among them are many  including eastern Long Island, Cape Hatteras in North Carolina, Palm Beach, Florida, and the region adjoining Lake Okeechobee in Florida  where the real estate boom has greatly increased the value of properties at risk. In many of these coastal communities, rather  than restricting development in the vulnerable beach areas, local governments are protecting ongoing development by building sea walls and restoring beaches, a costly and flawed approach.  



ii. the approach taken in this paper


The approach taken in this paper is to explore these issues by examining adaptations that have been made so far in the U.S. to the substantial climate changes that have already happened over the past half-century, as well as changes that are already inevitable because of greenhouse gas accumulations in the atmosphere up to this time. The extent and rapidity with which adaptations have been made to climate changes that have already happened or are inevitable should provide some guidance in assessing the likelihood of future adaptations and will help to identify those obstacles to adaptation that must be removed if future damages are to be limited.


The extent and rapidity with which adaptations have been made to climate changes that have already happened or are inevitable should provide some guidance in assessing the likelihood of future adaptations and will help to identify those obstacles to adaptation that must be removed if future damages are to be limited.


The U.S has already experienced considerable changes in temperature, precipitation, storm intensity and sea level due to climate change, and the effects of those changes have already been felt.


_ During the 20th century, average temperatures in U.S. increased by 1 degree Fahrenheit, with northern areas rising as much as 4 degrees Fahrenheit;

_ In the Northeast, temperatures in coastal regions have risen as much as 2 degrees Centigrade and precipitation has increased 20 percent in the last century;

_ In the Southeast, temperature and rainfall have increased. Sea level intrusion has destroyed coastal forests and many square miles of coastal land have been lost to erosion;

_ In Midwest, temperatures have risen 1 to 2 degrees Centigrade south to north and precipitation has increased, mostly in heavy storms resulting in more seasonal flooding;

_ In the Great Plains, there have been similar increases, but in the lee of the Rocky Mountains on the western plains, rainfall has diminished;

_ In the West, there have been rising temperatures, a larger fraction of winter precipitation falling as rain, a declining length of the snow season, reduced  snowpack, earlier snowmelt, and changed river flows. There has also been increasing drought in the Southwest.


In addition to these regional effects, the amount of precipitation in intense weather events has risen, sea levels have risen, hurricane intensity has increased, coastal erosion has increased (especially in Alaska, where a decline in seasonal sea ice along the coasts has exposed shorelines to much more storm erosion), and throughout the country, growing seasons and the ranges of some biotic organisms have shifted. Such changes as these over past decades are well-documented. Past increases in atmospheric concentrations of greenhouse gases have also already ensured as much climate change again in future decades, even if concentrations were miraculously stabilized overnight. Lags in the global climate system, due in part to the role of the oceans in absorbing and releasing heat, dictate that the effects of any rise in concentrations are not fully felt for decades. However, miracles will not occur. Realistically, few believe that concentrations can be stabilized at less than 550ppm, more than double pre-industrial levels. Many believe even that will be impossible. Any reasonable forecast would anticipate considerably more global warming and associated climate changes than have already been experienced.


Any reasonable forecast would anticipate considerably more global warming and associated climate changes than have already been experienced.


How have organizations in the United States responded to this reality? To sharpen the focus, I have looked primarily at organizations that are most likely to have undertaken adaptations because:

_ their operations, investments or missions are vulnerable to climate change;

_ they are making or planning long-term fixed investments or long-running programs which will inevitably feel the effects of climate change; and

_ they have the organizational capacity to forecast and plan.


In the public and quasi-private sector, the organizations that fit these criteria include, among others:

_ land and forest management agencies that will be exposed to ecological and hydrological changes;

_ flood control and disaster insurance agencies;

_ water-supply agencies in the West and Southwest that will feel the effects of drought and hydrological shifts;

_ public health and disease prevention agencies that will have to contend withchanges in disease vector ecology; and

_ transportation infrastructure agencies exposed to risks of sea level rise, storm surges and flooding.


For agencies in all these vulnerable sectors that are making long-term investments and commitments, our study investigated whether plans, designs, investment decisions, operational policies, budgets or staffing have been changed to reflect past or inevitable future climate change. If not, one must ask what obstacles have led to a lack of adaptation responses so far.


iii. obstacles to adaptation


Climate uncertainty


There are many reasons to doubt whether adaptation steps will be timely and efficient, even in the U.S. where the capabilities exist. Some of these doubts arise from the characteristics of the climate problem. Others arise from the tendency, exposed by behavioral economists, for inefficiencies in human and organizational behaviors. One of the most significant obstacles arises from the fact that most damages are incurred as the result of extreme weather events: unusual heat waves, droughts, floods, hurricanes and storm surges. These damages occur because most human and natural systems can tolerate climatic fluctuations within ranges, but tend to fail when conditions move outside those ranges. If a roof is built to withstand wind speeds of one hundred miles per hour, speeds below that rate may blow off a few shingles, but if speeds exceed one hundred miles per hour, the roof might well blow off, causing catastrophic damage to the structure. If a flood levee is designed to stop the flood likely to occur once in a hundred years, a flood greater than that will probably overtop the levee, causing severe flooding behind it. For such reasons, studies such as the Stern Review and work by economist William Nordhaus have found that climate damages rise very non-linearly with changes in weather variables.


Extreme events are infrequent by definition. Therefore, it is difficult to estimate their probability or frequency of occurrence, since there are so few observations in the historical record. Weather records may go back a century or so. How do we know what a once-in-five-hundred year flood might look like? In estimating the probability of extreme events occurring infrequently at the tails of the probability distribution, the particular underlying probability distribution that is assumed to represent the data becomes very important, since it will determine the likelihood of such extreme events when extrapolated beyond the range of existing data. Moreover, since there will be very few observations or data points representing these extremes, sampling error in fitting the distribution to the historical record will be large.


More fundamentally, those distributions estimated from historical data will be increasingly unrepresentative of future conditions as climate changes. The mean or average event might shift; so might the variance of the frequency distribution and the probability of extreme events. However, it is very difficult to judge when the probability of extreme events has changed. The degree of flooding in Iowa in June 2008 had not been experienced since 1851, if then. Does its occurrence signal that severe flooding has become more likely or does it merely represent a recurrence of a  very unlikely event? It might take decades and several occurrences to conclude with statistical certainty that what had been regarded as a once in a hundred year flood has become a once in a fifty-year flood.


Weather in most regions is notoriously variable. The noise to signal ratio in climate is large, making inferences about long-term changes difficult. Even trained scientists debated for decades whether climatic fluctuations exhibited any underlying trend, or whether observed changes merely reflected periodic fluctuations.11 It is much more difficult for untrained observers to detect relatively small trends amidst much larger short-term fluctuations. For the average person, this is all the more true because people tend to place undue weight on recent events, extrapolating short term movements into the future. Although the long-term trend might be slowly upwards, one or two cold years are sometimes taken as evidence that there is no underlying warming, for example.


Making matters worse, the variability of climate is typically much greater in specific locations than when averaged over a large area. Over a large area, many short term fluctuations cancel out but it is the local weather record on which people base their judgments. People judge climate change on the basis of conditions where they are. For example, the variability of temperature is considerably greater in Reno than for all of Nevada, and greater for Nevada than for the West as a whole. This makes it all the more difficult for people to discern trends. Moreover, because Americans are highly mobile, many people have not lived in the same locations for long periods of time, complicating the problem even further. Those are difficulties in interpreting the historical record. There are also difficulties in forecasting the future. Among climate scientists, there is still considerable uncertainty in projecting future climate change in particular regions because, although various global climate models might agree on global and broad continental-scale averages, they differ in their predictions for smaller regions. Regional climate models derived from the global circulation models also differ in their predictions. For example, planners in the West considering the construction of new dams to increase water storage capacity are struggling over predictions of future flows in the Colorado and other western rivers, which will depend on regional rainfall, snowpack, the timing of spring run-off, temperature and evaporation, and other factors, all of which are uncertain.12


As if these sources of uncertainty were not enough, climate change skeptics in and out of government have been deliberately sowing confusion and increasing public doubt about climate trends. For example, according to an inspector generals report at NASA, there was a sustained effort directed by political appointees to withhold information on climate change and to mute public statements by Dr. James Hansen, NASAs chief climate scientist. Our investigation, the report said, found that during the fall of 2004 through early 2006, the NASA Headquarters Office of Public Affairs managed the topic of climate change in a manner that reduced, marginalized or mischaracterized climate change science made available to the general public.13


In another well-reported episode, an official at the White House Council on Environmental Quality who was a former lobbyist for the American Petroleum Institute edited federal reports released in 2002 and 2003 on climate change to emphasize uncertainties and cast doubt on scientific findings. Also, major oil companies such as Exxon Mobil for years provided financial support to so-called climate skeptics, a small group of scientists who have persistently argued against the findings of the Intergovernmental Panel on Climate Change.


Moral hazard


Adaptation in some instances is inhibited by moral hazard issues. For example, governmental crop insurance and disaster relief programs have reduced the incentives for farmers, households, and businesses to take action to avoid weather damages. Between 1989 and 2007 indemnified losses insured by the Federal Crop Insurance Program increased from $1.2 to $3.8 billion, an average annual rate of increase of 6 percent per year. This insurance against crop loss is subsidized from the federal budget. During that period, the premium subsidy rose from $0.2 billion to $3.8 billion, an average annual rate of 17.6 percent per year.14 This subsidized insurance program provided strong incentives for farmers to take actions that increased their exposure to weather-related risks, which account for almost all losses. The incentives were further expanded by farm subsidies that raised farmers returns from agricultural operations whether or not crop damages occurred. Federal disaster relief program payments have also risen rapidly over past decades, both absolutely and as a percentage of assessed damages. In recent years these programs have covered roughly half of uninsured costs of weather-related disasters. Moreover, citizens, localities and state governments have come to expect federal disaster relief as an entitlement, accentuating the moral hazard problem.15


Organizational behavior


Organizations are inherently sluggish in responding to new conditions. In organizational decision-making, there is a strong status quo bias. The status quo is almost always the default option and deviations from it are almost always relatively small and incremental. More momentous changes occur only very infrequently, usually in response to major threats or pressures.16 Organizational actions and decisions are typically constrained by rules, routines,

procedures, formulae, and precedents. These are often codified, but even if not, they are enshrined as the way we do things and are usually difficult and costly to overturn and therefore may become badly out-of-date. Such routines are usually changed incrementally and reactively when existing routines prove to be unsuccessful.17 Interpretations of new experiences or evidence are made within frames of reference that are derived from past experience, and are also resistant to change. Moreover, adaptations to new conditions will usually be drawn from a repertoire of already known actions, competencies and strategies. Therefore, organizations change slowly and usually painfully.


Within these organizations, leaders and managers devote a very large fraction of their attention and the organizations resources to resolving immediate problems, leaving little with which to address longer-term problems like climate change. The perception that climate change is an issue that can be put off and doesnt demand immediate attention makes it likely that it will be put off and not get attention.


Behavioral economics


Behavioral economics has illuminated other characteristics of human decisionmaking under uncertainty that are likely to inhibit adaptation. Humans are myopic decision-makers, sharply discounting events in the farther future or past. In particular, people assign a relatively low priority to climate change because its effects are perceived to occur in the future, not the present:  People tend to underestimate cumulative probabilities when the probability of an event in a single period is low (i.e., the probability that the event will happen within x years is surprisingly high to most people). For example, people build or buy houses in fire-prone, flood and earthquake zones, even though the probability that an event will occur within their lifetimes is quite high;


 Humans exhibit strong anchoring to the status quo, tending to make only small adjustments away from it. Many people, for example, even refuse orders to evacuate when under immediate threat from natural disasters; _ People tend to resist and deny information that contradicts their value or ideological beliefs. An identifiable minority of climate skeptics continues to deny the scientific evidence and the conclusions of scientists regarding climate change.


Full report at:



Publications are produced using a print-on-demand system and printed on 100 percent recycled paper. For further information or inquiries, contact Jane Coppock, Editor of the Publication Series, at

Yale School of Forestry & Environmental Studies

publication series 205 Prospect Street New Haven, Connecticut 06511 USA