June 19, 2003 — Weather forecasts in the United States are provided through a partnership with NOAA and the private sector. The NOAA National Weather Service is the federal agency responsible for the collection, analysis and dissemination of weather forecast information, which is shared with the private sector who provides products to targeted users — such as TV, radio and newspapers. In turn, the largest customers of weather forecasts information are the 105 million U.S. households who consult the daily forecast on a regular basis. With an annual budget of about $1.4 billion, the average U.S. family pays about $13 a year for NOAA’s services. However, there is little if any reliable economic information indicating whether these services are worth this $13 a year to households, although it is widely assumed that the benefits far exceed the costs. Furthermore, there is even less information on the value to households of potential improvements to current weather forecasting services. As NOAA considers new programs and investments in new technologies, reliable information on the economic benefits of new expenditures is becoming even more critical.

Recognizing a need for better information on the economic value of weather forecasts, NOAA commissioned a study to assist in evaluating and quantifying — to the extent possible — the benefits to households of potential improvements in weather forecasting services, as well as how the public values current forecast services. Stratus Consulting in Boulder, Colo. was chosen to do the study because of its recognized expertise in valuing unique products like weather forecasts. Atmospheric Science Advisors LLC, in Silver Spring, Md., provided input on technical aspects of potential weather forecasting improvements.

The study focused on the values of day-to-day forecasts (In the future, hurricanes, tornadoes and snowstorms will also be addressed). Day-to-day forecasts (or ordinary weather forecasts) seemed like a reasonable place to start since 105 million U.S. households using them on a regular basis.

Estimating Values of Weather Forecasts—Public Goods and Non-Market Methods
In general, economists refer to day-to-day weather forecasts as “public goods,” similar to other public goods like national defense or environmental protection (Cornes and Sandler, 1996). Valuing public goods pose unique challenges because they aren’t traded in markets like automobiles or home appliances. This is due to two unique characteristics:

• First, weather forecasts are “non-excludable,” meaning that when the weather forecast is broadcast you can’t easily prevent an individual from getting access to that information. This makes it difficult or impossible to recoup the costs of the forecast by charging prices to users.
• Second, weather forecasts are “non-rival,” meaning that one person’s use of the forecast doesn’t diminish the value of that forecast to someone else. In this case, charging for the service would make consumers worse off because the marginal cost of an additional user is effectively zero.
• Given these characteristics, public goods like weather forecasts are not traded in traditional markets where people buy and sell goods (although the private sector sells tailored products to the media, which are then “bundled” with other features that contribute to media revenues or they sell highly specialized forecasts to specific users like farmers, railroads, etc).

Lacking data on prices and quantities, economists must rely on so-called “non-market valuation” methods, which have been successfully used in fields like environmental economics for several decades (Kopp, Pommerehne and Schwarz, 1997). These methods generally involve surveys of people to determine what public goods are worth to them. To elicit the value of weather forecasts, a survey instrument was developed through a series of focus groups, one-on-one interviews, a pilot study and external review by survey research experts. Data collection was implemented by individual, self-administered surveys conducted at survey centers in nine cities across the United States with a total of 381 participants.

Important Aspects of Weather Forecasts Identified
The survey first explored individuals uses of, perceptions of and sources for weather information. Precipitation and temperature were found to be the most important characteristics of weather forecasts and of current weather observations. Some weather information characteristics (such as dew point or barometric pressure) are infrequently used. Not surprising, wind strength and amount of precipitation was more important to those who work outside. For those who spend more leisure time outside, the chance of and amount of precipitation, extent of cloud cover, low temperature and how windy it will be were more important.

Values of Improved Weather Forecasts
Four attributes of weather forecasts were considered in the survey: the frequency of forecast updates, the accuracy of one-day forecasts, the accuracy of multi-day forecasts and the geographic detail of forecasts. Table 1 shows the baseline levels of these forecast attributes and levels of potential improvements considered by respondents (Potential improvements are the best feasible estimates made by technical experts, not "what-if" improvements).

Table 1
Attribute Levels for Survey

Two different non-market valuation approaches were used for estimating values of improved forecasts. First, stated choice methods were used (Louviere, Hensher and Swait, 2001) to examine values for potential changes in attributes of day-to-day weather forecast information. In this approach, two packages of improvements are “offered” using combinations of the attribute levels indicated in Table 1. The individual is asked which one she would prefer. These packages include a dollar amount indicating potential costs for getting those improved services. Ten different versions of the survey, each with six different choices between different packages of improvements, were used to explore individuals’ preferences for forecast improvements. Using statistical analysis, it is possible to determine how much marginal changes in each attribute are worth in dollar terms. A second method is to simply ask individuals their willingness to pay for a specific program that would improve weather forecast attributes.

Analysis indicated that improving the accuracy of one-day forecasts is valued most, followed by improving the accuracy of multi-day forecasts and geographic detail. Overall, individuals appear to have little value for increasing the frequency of weather forecast updates. While this holds for the day-to-day forecasts examined in this study, the frequency of updates is crucial in situations such as severe weather (e.g., tornadoes or floods).

As expected, values for improving weather forecasts are related to individuals’ characteristics — including income and education, how much time an individual spends outdoors on the job and how individuals use weather information in making behavioral decisions, such as what to do on the weekend or vacation.

Taken together, the value of a program to improve day-to-day weather forecasts to the maximum feasible quality for all four attributes was estimated at about $16 a year per household.

Values of Current Weather Forecast Services
A separate question was used to explore how much individuals thought current forecasts are worth. For this question, focus was on all services covered by weather forecasts — including severe weather, marine and aviation forecasts and other types of weather forecast information. More than 80 percent indicated they are willing to pay at least $32 for current services, with an estimated median household value of about $109 a year for current weather forecast services. The fact that households currently pay about $13 a year for these NOAA services indicates that there is a significant net benefit to the public from weather forecast information.

National Values for Weather Forecasts
Table 2 presents this study’s best estimates of annual per household values for improved and current weather forecasts. Based on U.S. Census Bureau estimates of about 105 million U.S. households, total national values for improving weather forecasts to the maximum levels proposed in the survey are estimated to be $1.73 billion per year. An annual national value of $11.4 billion was also calculated for current weather forecast services, which includes the value of all weather forecast information services from public and private sectors.

Table 2
Best Estimate Values for Improved and Current
Weather Forecast Services ($ 2001)

Benefits, Costs and the Contribution of Weather to the Economy
The aggregate national values for all current weather forecast services are estimated to be about $11.4 billion a year. With total federal spending on weather forecasting services of about $25 a year per household (including the FAA and DOD, along with NOAA), this study suggests a benefit-cost ratio of 4.4 to 1. A simple benefit-cost calculation indicates a net national benefit of $8.8 billion a year. However, this likely underestimates aggregate benefits because it may not include: (1) values generated outside of the household sector — like agriculture, transportation, or construction and (2) benefits to households in other countries that rely on meteorological data from the United States.

Another way to look at the contribution of weather forecasting to the U.S. economy is to keep in mind that estimated aggregate national benefits of improving day-to-day forecasts to the maximum feasible level are $1.73 a year. Using very “conservative” assumptions about costs, benefits, time horizons, discount rates, etc., the net present value of a program to modernize NOAA’s weather service, similar to the effort undertaken in the 1980s and ‘90s, is about $10 billion, or a nearly 10 percent annual rate of return on investment.

Next Steps
This work provides the first reliable estimates of the public’s value for improving weather forecasting. The work was informally reviewed recently by an independent panel of peer reviewers and by the National Academy of Sciences Board on Atmospheric Sciences and Climate. Both groups felt that this was a good approach for getting at the economic value of forecasts and that the results can be useful to policy makers, as well as in benefit-cost analysis. In particular, the NAS/BASC felt the work was an important step in implementing its recommendations in "Atmospheric Sciences: Entering the 21st Century (1999)" to better understand the economic and social impacts and benefits of atmospheric sciences. One suggestion by both groups was to continue this work and increase the sample sizes and implementation to obtain a truly nationally representative sample for estimating the values of current and improved weather forecasts.

NOAA is currently focusing on potential improvements in other services that may also generate significant benefits to households beyond the day-to-day forecasts considered in this study. For instance, work is underway to develop a survey for eliciting values for improved hurricane forecasting. Useful benefit assessments could also be undertaken for improvements for other severe weather forecasts and warnings, river level forecasts, large-scale weather systems and seasonal to interannual forecasts.

For additional reading on this study and the economics of the value of weather forecasts, see Jeffrey K. Lazo and Lauraine G. Chestnut, 2002, “Economic Value of Current and Improved Weather Forecasts in the U.S. Household Sector,” and, Atmospheric Science Advisors, “A Discussion of Selected Weather, Water, and Climate Forecasts and Services,” both available from Rodney Weiher, NOAA Chief Economist, at rodney.f.weiher@noaa.gov.

References
Cornes, R. and T. Sandler. 1996. The Theory of Externalities, Public Goods and Club Goods. 2nd edition. Cambridge University Press, New York.

Kopp, R., W. Pommerehn, and N. Schwarz, eds. 1997. Determining the Value of Non-Market Goods: Economic, Psychological, and Policy Relevant Aspects of Contingent Valuation Methods. Kluwer, Boston.

Louviere, J.J., D.A. Hensher and J.D. Swait. 2001. Stated Choice Methods: Analysis and Application. Cambridge University Press, Cambridge, U.K.

U.S. Census Bureau. 2000. America’s Families and Living Arrangements. Current Population Survey, Table 1. http://www.census.gov/prod/2001pubs/p20-537.pdf. Accessed March 27, 2002.

Relevant Web Sites
ECONOMIC STATISTICS FOR NOAA

AN ECONOMIC CASE FOR AN INTEGRATED OCEAN OBSERVING SYSTEM

WHAT IS THE “VALUE” OF THE BEACH?

ECONOMIC IMPACTS OF DROUGHT AND THE BENEFITS OF NOAA's DROUGHT
FORECASTING SERVICES

THE ECONOMIC IMPLICATIONS OF AN EL NIÑO

WEATHER IMPACT ON USA ECONOMY

NOAA National Weather Service

External Review of the Economic Value of Current and Improved Weather Forecasts Study

Economic Benefits of Weather and Climate Forecasts (Power Point Presentation)

Media Contact:
Greg Hernandez, NOAA, (202) 482-3091