Contingent Valuation Method (CVM)
Given the wide range of users of Landsat imagery and various uses the imagery is put towards, significant economic benefits are likely to be generated from its use. However, determining these benefits can be a difficult task, partly because there is no market price to reflect the value of the imagery to society. Landsat imagery has characteristics of a public good, meaning the socially optimal level of provision through private markets is not likely. Even the previous price did not accurately reflect economic benefits because it was administratively set. To measure these economic benefits accurately, consumer surplus is the appropriate measure. This is the standard measure of benefits in benefit-cost analysis and the Office of Management and Budget also recommends using it, stating, “When it can be determined, consumer surplus provides the best measure of the total benefit to society from a government program or project.” Economists use a range of methods to monetize the economic benefits provided by goods and services that are not traded in markets. When there is no price, or there is little or no market data available on the benefits to users, a stated preference or intended behavior technique known as the contingent valuation method (CVM) is commonly used. CVM is a survey-based approach used to estimate the economic benefits individuals receive from a nonmarket good or service.
Landsat User Surveys
As part of determining the value of Landsat imagery to users, we asked a dichotomous-choice format question in the 2009 and 2012 surveys (for more information about the surveys, click here) to measure the economic benefits from Landsat imagery. The user was asked to decide whether a Landsat-equivalent scene is worth the cost specified in the question. Since the results from the 2012 survey were generalizable to a population and the results from the 2009 survey were not, this discussion focuses on the 2012 survey. However, a very similar approach was used in both surveys. The specific question asked in 2012 was:
“At the moment, current Landsat 5 imagery is not available (expected to be available again in spring of 2012) and you may have already obtained imagery elsewhere to replace Landsat 5. If both Landsat 5 and 7 became permanently inoperable before the next Landsat satellite is operational (scheduled to launch in early 2013), you may have to obtain imagery elsewhere again. Assume that you are restricted to your current project or agency budget level and that the money to pay this cost would have to come out of your existing budget. If such a break in continuity did occur and you had to pay for imagery that was equivalent to the Landsat standard product typically available (which assumes both Landsat 5 and 7 imagery are available), would you pay $X for one scene covering the area equivalent to a Landsat scene?”
Respondents were instructed to answer Yes or No. The range of bid amounts was determined from the 2009 survey of Landsat users. This question includes an explicit budget constraint (“Assume you are restricted to your current project or agency budget level…”) and a reminder that the funds to pay the higher cost would have to come out of this fixed budget. This follows the recommendation of the National Oceanic and Atmospheric Administration panel on contingent valuation that budget reminders are to be included in CVM survey questions. Recognition of budget constraints is important to be consistent with consumer behavior and demand theory.
Using a follow-up CVM question allows for improved inference of economic benefits. If the respondent answered Yes to the first question, then a second question asked if they would pay a higher amount. If the respondent answered No to the first question, then a second question asked if they would pay a lower amount. The response to these two questions leads to a series of Yes/Yes, Yes/No, No/Yes, and No/No answers, providing the data necessary to calculate a “double-bounded” estimate of economic benefits. The responses are regressed on the bid amount, demographic characteristics of the respondent, and other relevant variables. The results of this regression model can then be used to monetize the average and median economic benefits provided by one Landsat scene. In the previous survey conducted in 2009, a follow-up CVM question was also included; however, due to difficulty in interpreting the results, only the results from the original question were used, resulting in a “single-bounded” estimate. For the 2012 survey, we used a new bid design for the follow-up question that was informed by the 2009 survey.
2012 Survey Results
The main set of results reported here are from the double-bounded CVM question using the new bid design, because these are the most precise results. We report results for four groups of users: (1) U.S. established users, (2) U.S. new/returning users, (3) international established users, and (4) international new/returning users. Currently, U.S. users download the overwhelming majority of scenes from EROS, so it was important to understand any differences between the benefits provided to U.S. versus international users. We also hypothesized that established users would report greater benefits from using Landsat imagery than new/returning users, based on their consistent use of the imagery over time.
Results from this analysis show that the median value of the economic benefits, or consumer surplus, obtained from Landsat imagery was $182 per scene (90% confidence interval (CI) = $157-$207) for U.S. established users and $49 per scene (90% CI = $42-$55) for U.S. new/returning users (table 1). This is not the value of the scene to the typical user but the value where half (50 percent) of the sampled users would purchase a scene equivalent to a Landsat scene. The median can also be thought of as the amount where half the Landsat users registered with EROS would not purchase a scene and thus would cease to receive economic benefits from the imagery. The mean consumer surplus or average value of the economic benefits was $912 per scene (lower bound (LB) = $829) for U.S. established users and $367 per scene (LB = $341) for U.S. new/returning users (table 1; averages were weighted and truncated, see appendix 2 in the full report for more information). The purpose of the confidence interval and the lower bound is to communicate some of the variation associated with the median and average point estimates. The main conclusion from both, however, is that point estimates are relatively precise. Both the median and average were substantially less for new/returning U.S. and international users than for established users. This would be expected, as the new and returning group of users was motivated to begin using, or return to using, Landsat imagery as a result of the free and open data policy. The average was much higher than the median for all groups of users because there is a small, but significant, group of users that values Landsat imagery very highly. This may be due to the nature of the respondents who are generally technically oriented, professional, and knowledgeable about the good they were asked to value.
To calculate the annual aggregate value of Landsat imagery, two pieces of information are necessary: (1) the number of scenes obtained by each of the four groups of users (U.S. and international established and new/returning users) and (2) the average economic benefit per scene for each group. The latter was calculated using the contingent valuation method outlined above. The former uses data provided by EROS and from the survey. EROS keeps records of the number of scenes downloaded each year to U.S. and international users but does not have information on which users are established and which are new or returning users. Using information from the survey on the average number of scenes that each of these groups obtains annually from EROS, the proportion of scenes obtained by each group was calculated. For U.S. users, established users were estimated to have obtained 71% of the scenes annually and new/returning users obtained 29%. For international users, established users were estimated to have obtained 60% of the scenes annually and new/returning users obtained 40%. By applying these proportions to the total number of scenes distributed by EROS to U.S. and international users in 2011 (the last full calendar year before the survey was administered), an estimate of the number of scenes each group obtained in 2011 was attained. The annual value of Landsat is the average value per scene for each group multiplied by the total number of scenes each group obtained in 2011 (table 2). The annual economic benefit from Landsat imagery obtained from EROS in 2011 was just over $1.79 billion (LB = $1.64 billion) for U.S. users and almost $400 million (LB = $363 million) for international users, resulting in a total annual economic benefit of $2.19 billion (LB = $2 billion).
Interpretation of CVM Results
The information presented in this analysis may raise the question of whether users should be charged to obtain Landsat imagery. As mentioned previously, similar to other data and information sources, Landsat imagery has characteristics of a public good. Specifically, the imagery is nonrival in consumption, meaning more than one person can use the same imagery at the same time, and once the imagery is made publicly available, the additional, or marginal, cost of allowing one more person to use it is zero. The inability of the private sector to supply public goods efficiently is a type of market failure, and the Federal Government plays an important role in the provision of data and information sources that are not efficiently provided by the private sector. Assuming Landsat imagery continues to be provided by the public sector, economic analysis can be used to determine the efficient price to charge users for the imagery. The relation between the economic benefits society receives from the use of Landsat imagery at different price points can be shown graphically. An illustrative demand curve for Landsat imagery is shown in figure 1, with the quantity of scenes downloaded on the horizontal axis and the price per scene on the vertical axis.
The demand curve for Landsat imagery slopes downward, following the law of demand. As the price per scene increases, the quantity of scenes demanded decreases and vice versa. There is considerable evidence to support this downward sloping demand curve for Landsat imagery. First, when prices reached $4,400 a scene during the era of privatization, many users were priced out of the market and switched to other imagery. Second, since the imagery has been available at no cost, the number of users and number of scenes downloaded has increased substantially. Finally, the CVM analysis presented here confirms what is expected: at higher bid amounts, there is a lower probability that survey respondents would pay for the imagery. At different price points, both the number of imagery users and the number of scenes downloaded would be expected to change. Economic benefits, or consumer surplus, can be illustrated graphically as the area under the demand curve and above the price paid for a particular good. At a price of $0 per scene, the entire area under the demand curve reflects the economic benefits received from the use of Landsat imagery, as shown in figure 2.
Figure 3 shows the effects of charging a positive price, $X*, for the use of Landsat imagery. At this price, a quantity of Q* scenes would be downloaded. The economic benefits that users would receive are illustrated by the shaded area with diagonal stripes. The loss in economic benefits to users who continue to use Landsat imagery but now pay $X* per scene is illustrated by the shaded area with horizontal stripes. This same amount would be transferred to the government as revenue. Finally, the shaded area with vertical stripes represents a loss in economic benefits due to users who are not willing to pay $X* per scene exiting the market for Landsat imagery. This area is a combination of remaining users downloading fewer scenes and other users completely exiting the market because they value the image at less than $X* per scene. Because this loss in surplus accrues to no one as a gain, it is referred to in economics as deadweight loss. Charging any positive price for a nonrival good is economically inefficient; it results in under consumption of the good and a net loss of economic benefits to society.
The CVM analysis presented in this section shows that, though there is a small group of users who would pay a lot for Landsat imagery, most users are not willing, or able, to pay very much. More than two-thirds (68%) of users responded No to both CVM questions (and thus both bid amounts) they received. Charging a positive price for the use of imagery would restrict access with no associated net gain to society. It would prevent some individuals from receiving benefits from the use of the imagery either by causing them to exit the market or by causing them to obtain fewer scenes than they would otherwise, while the Federal Government would incur no additional cost in letting these users have access to the imagery. As shown in figure 3, charging a positive price for a nonrival good such as Landsat imagery results in a deadweight loss to society; even charging a minimal price would result in considerable economic losses. The economically efficient price to charge users is zero.
Further, there is good reason to believe that charging a positive price for the use of imagery would result in even greater economic losses than those shown in figure 3. Charging for the imagery would hinder innovation resulting from its use. When users paid a price per scene, they downloaded fewer scenes than they do currently, and as a result, were constrained in the uses they could put the imagery towards. Once it became available at no cost, users could download as many scenes as they wanted and use the imagery in new applications, some of which never would have existed if the imagery had not become available at no cost. These uses generate societal benefits that would be reduced in the absence of a free and open data policy. In addition, the discussion thus far has focused on direct users of Landsat imagery. There are many downstream users of Landsat imagery and imagery-derived products. Charging a positive price would also reduce the benefits obtained from these downstream uses.
Arrow, K., Solow, R., Portney, P.R., Leamer, E.E., Radner, R., and Schuman, H., 1993, Report of the NOAA panel on contingent valuation: Wash., D.C., National Oceanic and Atmospheric Administration, U.S. Department of Commerce.
Cameron, T., and Quiggin, J., 1994, Estimation using contingent valuation data from a dichotomous choice with follow-up questionnaire: Journal of Environmental Economics, v. 27L, p. 218-234.
Champ, P.A., Boyle, K.J., and Brown, T.C. (Eds.), 2003, A primer on nonmarket valuation (The economics of non-market goods and resources series, Bateman, I.J. (Ed.)): Norwell, Mass., Kluwer Academic Publishers, 577 p.
Krugman, P., and Wells, R., 2009, Microeconomics (2nd ed.): New York, Worth Publishers, 567 p.
Miller, H.M., Sexton, N.R., Koontz, Lynne, Loomis, John, Koontz, S.R., and Hermans, Caroline, 2011, The users, uses, and value of Landsat and other moderate-resolution imagery in the United States—Executive report: U.S. Geological Survey Open-File Report 2011-1031, 43 p.
Office of Management and Budget, 1992, Guidelines and discount rates for benefit-cost analysis of federal programs, Circular A-94, Revised Transmittal Memo #64: Wash., D.C., Executive Office of the President.
Sassone, P., and Schaffer, W., 1978, Cost benefit analysis—A handbook: New York, NY, Academic Press.
U.S. Environmental Protection Agency, 2000, Guidelines for preparing economic analyses, EPA 240-R-00-003: Wash., D.C., U.S. Environmental Protection Agency.
U.S. Water Resources Council, 1983, Economic and environmental principles and guidelines for water and related land resources implementation studies: Wash., D.C., U.S. Government Printing Office.
Whitehead, J., 2002, Incentive incompatibility and starting point bias in iterative valuation questions: Land Economics, v. 78, no. 2, p. 285-297.