7

Evaluation of Economic Modeling

OVERVIEW OF METHODS AND ANALYSIS

The economic analysis in the updated site-specific risk assessment (uSSRA) of the proposed National Bio- and Agro-Defense Facility (NBAF) is more comprehensive than other economic analyses of foot-and-mouth disease (FMD) outbreaks (Ekboir, 1999; Paarlberg et al., 2002, 2008; Pendell et al., 2007; Hagerman et al., 2012). The uSSRA relies on three components to predict the economic consequences of an FMD epidemic: a partial equilibrium model, a regional model, and accounting of government non-indemnification costs. A partial equilibrium model of the U.S. agricultural sector describes supply and demand for U.S. livestock products, livestock, and major feed crops in global markets. A regional model captures the changes in expenditure in the 7-state region surrounding the NBAF. The models determine the impact of depopulating livestock during an outbreak, of reductions in U.S. consumer demand for livestock products, and of reduced exports of livestock products on market prices, quantities, economic welfare, and expenditure. The change in economic welfare is determined by adding the changes in returns to producer capital and management, consumer surplus, indemnification costs, changes in regional non-agricultural expenditure, and government non-indemnification costs, including surveillance, destruction, and mitigation. Compared with the 2010 SSRA, the updated analysis relies on an updated baseline with solutions for 40 quarters versus 20 quarters. Revisions in the epidemic modeling that extend the magnitude and duration of an outbreak are captured in revisions of the economic results. The uSSRA’s use of a dynamic equilibrium model and a



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 63
7 Evaluation of Economic Modeling OVERVIEW OF METHODS AND ANALYSIS The economic analysis in the updated site-specific risk assessment (uSSRA) of the proposed National Bio- and Agro-Defense Facility (NBAF) is more comprehensive than other economic analyses of foot-and-mouth disease (FMD) outbreaks (Ekboir, 1999; Paarlberg et al., 2002, 2008; Pendell et al., 2007; Hagerman et al., 2012). The uSSRA relies on three components to predict the economic consequences of an FMD epidemic: a partial equilibrium model, a regional model, and accounting of government non-indemnification costs. A partial equilibrium model of the U.S. agricul- tural sector describes supply and demand for U.S. livestock products, live- stock, and major feed crops in global markets. A regional model captures the changes in expenditure in the 7-state region surrounding the NBAF. The models determine the impact of depopulating livestock during an outbreak, of reductions in U.S. consumer demand for livestock products, and of re- duced exports of livestock products on market prices, quantities, economic welfare, and expenditure. The change in economic welfare is determined by adding the changes in returns to producer capital and management, consumer surplus, indemnification costs, changes in regional non-agricul- tural expenditure, and government non-indemnification costs, including surveillance, destruction, and mitigation. Compared with the 2010 SSRA, the updated analysis relies on an updated baseline with solutions for 40 quarters versus 20 quarters. Revisions in the epidemic modeling that extend the magnitude and duration of an outbreak are captured in revisions of the economic results. The uSSRA’s use of a dynamic equilibrium model and a 63

OCR for page 63
64 NBAF UPDATED SITE-SPECIFIC RISK ASSESSMENT regional model are appropriate, but the committee has concerns about the analysis for reasons noted below. INACCURATE DESCRIPTIONS OF METHODS AND ANALYSIS Descriptions of the model and scenarios in the uSSRA are sometimes inaccurate. The partial equilibrium model is described as temporal and spatial; the latter is inaccurate. The model is temporal in that it solves dy- namically to capture the lags inherent in animal and crop production. The model is not spatial inasmuch as it treats the United States as a point in space, so the uSSRA text incorrectly implies that the model handles regional differences in production and consumption and hence regional trade, which the model is not capable of handling. The uSSRA also states that observed values are used for the 2009–2018 model baseline (p. 546), but observed values are only possible for 2009 and 2010 and not future timeframes. For 2011–2018, the model baseline is the annual U.S. Department of Agriculture (USDA) baseline from Febru- ary 2011 converted into quarterly values. That means that there is more fluctuation in the 2009 and 2010 results because the USDA baseline tends to smooth out. The FMD supply shocks are introduced into the more vari- able baseline values. The inclusion of farm programs based on the 1996 and 2002 farm bills is mentioned on p. 545 of the uSSRA, but this version of the model incorporates the 2008 farm bill and the Average Crop Revenue Election (ACRE) program. Given the baseline crop prices, the only change in government payments to crop producers is via ACRE unless reduced feed consumption causes crop prices to fall enough to trigger traditional com- modity program payments. Thus, the timeframe of 2009–2018 matters for the results because changes in U.S. government payments will not buffer losses in returns to crop growers except for the small share of production enrolled in ACRE. In the uSSRA, there is a tendency for the text to refer to cattle, and this leaves the situation for other species vague. For example, it is unclear whether the vaccinate-to-live strategy applies to swine and sheep as well as to cattle. The uSSRA should have described the domestic demand shocks as percentage reductions in domestic demand rather than as consumers dropping meat consumption, because consumers could reduce consumption without completely avoiding red meats. INSUFFICIENT INFORMATION PROVIDED TO VERIFY RESULTS The quality of the estimates of economic impacts depends on input from the North American Animal Disease Spread Model, correct operation of the economic models, and decisions on scenario design. The magnitudes of livestock depopulation are estimated in the epidemic modeling. Those

OCR for page 63
65 EVALUATION OF ECONOMIC MODELING values are converted into supply reductions and combined with demand shocks, including U.S. domestic consumer behavior and export reductions. The text and tables in the uSSRA are insufficient for the reader to replicate and evaluate the economic results. At times, the discussion in the uSSRA does not correspond with infor- mation provided in the tables. The discussion of how supply shocks are estimated is confusing, and the information in the tables cannot be used to determine whether the supply shocks are implemented correctly. A more detailed table to illustrate the magnitudes of supply reductions for a multi- quarter outbreak would have been useful. The demand shocks as described in the text and as reported in Table 7.3.1-9 appear to be in conflict, prob- ably because of the wording of column headings (p. 563 of the uSSRA). The text indicates that domestic demand begins to recover in the quarters immediately after the end of an outbreak, but the table shows 10% demand reductions in the p50/p50 case for the end of the outbreak + four quarters with recovery beginning in the 5th quarter after the outbreak ends. PARTIAL EQUILIBRIUM MODEL ANALYSIS The partial equilibrium model includes livestock products, animals, and major crops at a national level. In response to supply disruptions and reductions in domestic and export demands, the partial equilibrium model determines the quarterly time path of changes in market prices, quantities, and measures of economic welfare. Government indemnification, surveil- lance, mitigation, and control costs are also calculated. The uSSRA uses the partial equilibrium model, and this is a proper approach. The validity of the analysis depends on the quality and validity of the economic shocks and how the model is used. The committee finds that the upper ranges for the supply shocks, the outbreak duration, and other aspects of scenario de- sign, which were derived from the epidemic modeling, are underestimated. The committee therefore concludes that the magnitudes of the estimated economic impacts are also underestimated. In responding to an FMD outbreak, the vaccinate-to-live decision for outbreaks consisting of 180 days or more is important for the results. Vac- cinated animals remain in the supply chain, and knowledge that vaccinated animals are in the meat supply could affect consumer response. The text is unclear about how vaccinate-to-live is reflected in the modeling: it is un- clear whether those animals are excluded from the supply shocks. If so, it is unclear whether the supply shocks go to zero for quarters 3 and beyond or whether animals are still waiting destruction; the latter could affect contin- ued supply reductions. The text does not explain that the greater assumed demand reductions for longer outbreaks partly incorporate a consumer response to having vaccinated animals in the supply chain. The change in producer welfare is described as producer surplus, but it

OCR for page 63
66 NBAF UPDATED SITE-SPECIFIC RISK ASSESSMENT is actually the return to capital and management. Although they are similar concepts, there are slight differences in these measures. In the partial equi- librium model, crop producers are separated from land owners, so all crop producers are modeled as tenants to isolate the changes in land rent. The change in returns to land owners should have been included in the reported uSSRA results. REGIONAL MODEL ANALYSIS Regional modeling determines non-agricultural changes in economic activity in the 7 states experiencing an outbreak. The regional analysis is an important component because it captures economic impacts on sectors within the 7-state region that are excluded from the partial equilibrium model. Although the regional non-agricultural costs are not large—$40 mil- lion to $6 billion in the context of the total economic impact of $16–140 billion—it is important to include such costs. These costs are concentrated in a 7-state region and concentrated on a small set of industries within that region. Because the regional model and modeling approach differ from the partial equilibrium model, care must be exercised when using these in the same analysis. The Regional Input-Output Modeling System (RIMS II) developed by the Bureau of Economic Analysis is well known and has been available for some time. Although there is extensive discussion of the regional model in general, decisions were made about multipliers (such as regional aggregation and leakages), and the multipliers used in the regional model were not included in the uSSRA but are necessary for evaluating the regional results. Table 7.3.2-2 of the uSSRA reports expenditures for travel by state. The effect of an outbreak on travel is a major component of the regional analysis. The use of changes in expenditure as a measure of regional non- agricultural welfare effects as suggested by the table on p. 568 is a concern. A change in expenditure is not a measure of a change in economic welfare unless utility is held constant. But Table 7.4.1-1 of the uSSRA shows that changes in regional non-agricultural expenditure are added to the changes in producer returns to capital and management, consumer surplus, indem- nification, and government non-indemnification costs to give the change in total economic welfare. The change in expenditure is correctly described in the text as the change in economic activity. Changes in expenditure exceed changes in economic welfare because the change in expenditure does not account for adjustment in variable costs. An alternative closer to theoretically correct measures of economic welfare—such as the return to non-traded, quasi-fixed inputs (value-added)—would have been a better choice if available.

OCR for page 63
67 EVALUATION OF ECONOMIC MODELING NON-INDEMNIFICATION COSTS The values for government costs used in the 2010 SSRA are adjusted for inflation and are used to determine government non-indemnification costs. There is extensive reporting of depopulation and vaccination costs in Section 6, but it is not clear why that information is not used to model economic impact in Section 7 of the uSSRA. Government costs that would have been incurred regardless should be excluded, and only additional costs incurred by government should be included; otherwise, the estimated government non-indemnification costs that are reported are higher than the actual additional costs incurred. SUMMARY The economic modeling approaches used in the uSSRA are frameworks that, if used properly, can address the economic impacts of an FMD out- break that results from a breach of containment at the NBAF. The uSSRA attempts a more comprehensive analysis than existing analyses. Neverthe- less, there remain concerns about the uSSRA. The information provided in the text and tables is insufficient to determine whether the analysis was done correctly. There are concerns about aggregating the results of differing modeling approaches into an aggregate economic impact. The economic consequences are generated from results obtained in the epidemic model- ing. The committee believes those depopulation and duration estimates are underestimated, and consequently the supply reductions and the demand re- ductions used as inputs into the economic analysis are also underestimated. REFERENCES Ekboir, J.M. 1999. Potential Impact of Foot-and-Mouth Disease in California: The Role and Contribution of Animal Health Surveillance and Monitoring Services. Agricultural Issues Center, Division of Agriculture and Natural Resource, University of California, September 1999. Hagerman, A.D., B.A. McCarl, T.E. Carpenter, M.P. Ward, and J. O’Brien. 2012. Emergency vaccination to control foot-and-mouth disease: Implications of its inclusion as a U.S. policy option. Appl Econ Perspect Pol 34(1):119-146. Paarlberg, P.L., J.G. Lee, and A.H. Seitzinger. 2002. Potential revenue impacts of a foot and mouth disease outbreak in the United States. J. Am. Vet. Med. Assoc. 220(7):988-992. Paarlberg, P.L., A. Hillberg Seitzinger, J.G. Lee, and K.H. Mathews Jr. 2008. Economic Impacts of Foreign Animal Disease. Economic Research Report ERR-57. Economic Research Service, U.S. Department of Agriculture: Washington, DC. Pendell, D.L., T.C. Schroeder, J. Leatherman, and G.S. Alward. 2007. The economic impacts of a foot-and-mouth disease outbreak: A regional analysis. J Agr Appl Econ 39:19-33.

OCR for page 63