Advancing Computable General Equilibrium Analysis (CGE) of the Economic Consequences of Terrorism

Transcription

1 CREATE Research Archive Research Project Summaries 2012 Advancing Computable General Equilibrium Analysis (CGE) of the Economic Consequences of Terrorism Peter B. Dixon Monash University, James Giesecke Monash University, Maureen T. Rimmer Monash University George Verikios Monash University Glyn Wittwer Monash University Follow this and additional works at: Part of the Behavioral Economics Commons, and the Econometrics Commons Recommended Citation Dixon, Peter B.; Giesecke, James; Rimmer, Maureen T.; Verikios, George; and Wittwer, Glyn, "Advancing Computable General Equilibrium Analysis (CGE) of the Economic Consequences of Terrorism" (2012). Research Project Summaries. Paper This Article is brought to you for free and open access by CREATE Research Archive. It has been accepted for inclusion in Research Project Summaries by an authorized administrator of CREATE Research Archive. For more information, please contact

2 National Center for Risk and Economic Analysis of Terrorism Events University of Southern California Los Angeles, California Advancing Computable General Equilibrium Analysis (CGE) of the Economic Consequences of Terrorism October 2011 to September 2012 Peter Dixon, James Giesecke, Maureen Rimmer, George Verikios and Glyn Wittwer Monash University, Centre of Policy Studies "This research was supported by the United States Department of Homeland Security through the National Center for Risk and Economic Analysis of Terrorism Events (CREATE) under Cooperative Agreement 2007-ST-061-RE0001. However, any opinions, findings, and conclusions or recommendations in this document are those of the authors and do not necessarily reflect views of the United States Department of Homeland Security or the University of Southern California." Cooperative Agreement No ST-061-RE0001 Department of Homeland Security September 28, 2012

3 ABOUT CREATE The National Center for Risk and Economic Analysis of Terrorism Events (CREATE) was the first university-based Center of Excellence (COE) funded by University Programs of the Science and Technology (S&T) Directorate of the Department of Homeland Security (DHS). CREATE started operations in March of This annual report covers the seventh year of CREATE funding from October 2010 to September 2011, the first year under Cooperative Agreement 2010-ST-061-RE0001 from DHS. While the text of this report focuses on the seventh year, all data tables, publications, lists of participants, students, and presentations and events are cumulative from the inception of CREATE. CREATE s research mission is to develop advanced models and tools for risk assessment, economic assessment, and risk management to counter terrorism. CREATE accomplishes this mission through an integrated program of research, education, and outreach, spanning the disciplines of economics, psychology, political science, industrial and systems engineering and information science. CREATE develops models, analytical tools, methodologies and software, and tests these tools in case analyses, representing critical homeland security investment and policy decisions. Due to the cross-cutting nature of research in risk, economics, and risk management, CREATE serves the need of many client agencies at the DHS, including the Transportation Security Agency, Customs and Border Protection, Immigration and Customs Enforcement, FEMA and the US Coast Guard. In addition, CREATE has developed relationships with clients in the Offices of National Protection and Programs, Intelligence and Analysis, General Council, Health Affairs, and Domestic Nuclear Detection. Using a mix of fundamental and applied research, CREATE faculty and students take both the long-term view of how to reduce terrorism risk through fundamental research and the medium-term view of how to improve the costeffectiveness of counter-terrorism policies and investments through applied research. Please visit for more information. Page 2 of 10

4 Advancing Computable General Equilibrium Analysis (CGE) of the Economic Consequences of Terrorism Peter Dixon, James Giesecke, Maureen Rimmer, George Verikios and Glyn Wittwer Monash University, Centre of Policy Studies 1. Executive Summary Publications and Reports Presentations Outreach Executive Summary CREATE has evaluated several modeling approaches to estimating the national and regional consequences of terrorist attacks. In 2008 we utilized the leading regional/national econometric package, REMI, to analyze the economic impacts of a shutdown of the US borders to trade, travel, and tourism in the face of a public health threat or terrorist attack. Unfortunately, we found REMI to have significant shortcomings for cases of major shocks to the economy that involve issues of interregional and international competition. Progress has been made by CREATE researchers in improving input-output models in the areas of interregional interactions (NIEMO) and some incorporation of resilience considerations (primarily Flex-NIEMO). However, I-O remains limited in many applications due to its inherent linearity, lack of behavioral content, and inability to represent the detailed workings of markets. On the other hand computable general equilibrium (CGE) modeling has been successfully applied to a number of terrorist scenarios and has the capability to be applied to a broader range of applications than any of the alternatives. Peter Dixon at the Centre of Policy Studies (CoPS) at Monash University, Australia, has pioneered many advances in CGE modeling (Dixon et al., 1982; 2002). More recently, he and his research team have performed studies for the International Trade Commission and for DHS (Dixon and Rimmer, 2009, 2010). The latest generation of his CGE models has continued this path-breaking research theme by adding dynamic elements, so that models can be run on a quarterly rather than annual basis. At the same time, CREATE has pioneered the inclusion of resilience into CGE models and has successfully constructed and applied bottom-up regional models to terrorist attacks. However, the CREATE CGE models are annual and static. Much has been accomplished by a collaboration between CREATE and CoPS. This is facilitated by interactive research at each site and through visits by Peter Dixon, Maureen Rimmer, James Giesecke and George Verikios of CoPS to CREATE. In the year ended September 2012 the CoPS group continued its CREATE-related work in economy-wide modelling in three stands. First, the group continued its fundamental research in regional modelling, centred on the creation and refinement of USAGE-R51. This is the first bottom-up interregional CGE model of the U.S. Potentially it specifies each of the 50 states of the U.S. plus the District of Columbia, at the 500 industry level, and treats each state as a separate economy linked to the other 50 U.S. regions via: trade in goods and services; migration of labor; financial flows; and the use of a single currency. The model is designed flexibly so that it can be run at different levels of regional and industrial aggregation. Since September 2011, several test runs have been carried out and extensive work has been done on the database. Page 3 of 10

5 The database of USAGE-R51 has greater sectoral detail than that of the national USAGE CGE database. The present master database of USAGE-R51 contains 514 sectors in each of the 51 regions. Electricity has been split into eight types of generation plus distribution. Agriculture has been split into crop and livestock activities. The electricity split means that USAGE-R51 could, with further development, become a valuable tool for examining disaster-related interruptions to power supplies in the United States. The agriculture split aligns with data available by state and will enhance the capability of USAGE-R51 in examining for example, food security issues. Source Table 1: Data sources used in preparation of USAGE-R51 database USAGE national CGE database Bureau of Economic Analysis national accounts data (60+ sectors) Task Update to 2005 Used twice: (1) national targets for update to 2005 (2) State shares of national activity control totals 2007 census Annual payroll bills, number employed to estimate regional shares of national activity for 445 out of 514 sectors USDA data U.S. Energy Information Administration U.S. Department of Health and Human Services Medicaid and Medicare data National Center for Education Statistics Used twice: (1) national targets for split of initial CGE database (2) State shares of national activity by crop or livestock activity Used twice: (1) split of initial CGE database into electricity generation by fuel type (2) State shares of national activity by type of generation Public health spending shares by state Split SLGOther into state production shares Government expenditure shares State shares of SLGovEduc International merchandise trade by port used to estimate international state export and import shares Table 1 outlines the data sources used in preparing USAGE-R51. The task of preparing the master database of USAGE-R51 is made reproducible and mechanical by the use of computing programs. This means that as better data surface, such as for example, more disaggregated international trade data or more detailed household expenditure data, new Page 4 of 10

6 information on state shares can be added to the input files of the programs. Rerunning the master database creation jobs is relatively quick and easy. The master database of USAGE-R51 contains many more regions and sectors than a particular study will ever require. To provide a database that is both computable and for presentational purposes, tractable, the master database is aggregated in regions and sectors of little interest in particular study while retaining full disaggregation of sectors and state of interest in the study. It may be the case that undertaking a particular study leads to the discovery of deficiencies in the USAGE-R51 database. If this is so, amendments can be made to state shares or even to the national CGE database to rectify the deficiency. The database creation programs can then be run again, including the improvements. The second strand of CoPS collaborative work with CREATE in 2012 was the application of economy-wide models to issues in security/terrorism. This research has encompassed a range of terrorism and catastrophic events (radiological dispersion device attack, chlorine gas attack, and pandemic influenza). Giesecke et al. (2012) modelled three avenues via which terrorism-induced changes in risk perception can affect a regional economy: (i) required rates of return on investment; (ii) willingness to pay for goods associated with the region; and (iii) compensating wage differentials. Their modelling was undertaken within a comparative-static framework. However the paper raised the possibility that in future research a dynamic framework might be better equipped to track the typical path of peak and decay exhibited by fearinduced behavioral effects. The structure of existing regional dynamic CGE models can readily accommodate shocks describing (i) and (ii). However they are less well suited to (iii). Convincing dynamic modelling of (iii) requires an inter-regional migration theory that embodies: (a) imperfect interregional labour mobility; (b) gradual adjustment in inter-regional wage differentials via inter-regional mobility; and (c) modelling of gross (not net) inter-regional population flows. In preparing dynamic regional CGE models to better represent the dynamic path of fear-induced movements in wage requirements, we have developed an inter-regional migration theory that addresses (a) (c). We have implemented this in two dynamic CGE models: (i) a dynamic version of the single-region model described in Giesecke (2011) and employed in Giesecke et al. (2012) to examine an RDD scenario; and (ii) a multi-regional dynamic model of a national economy. This work has been presented at two conferences, see section 3. Verikios et al. (2011) analyses the global economic effects of two influenza pandemics that represent extremes along the virulence-infectiousness continuum of possible pandemics: a high virulence-low infectiousness event and a low virulence-high infectiousness event. They do this by applying results from a susceptible-infectedrecovered epidemiological model to a detailed, quarterly computable general equilibrium model. Their findings indicate that global economic activity is more strongly affected by a pandemic with high infection rates rather than high virulence rates, all else being equal. At the regional level, regions with a higher degree of economic integration with the world economy are affected more strongly than less integrated regions. Page 5 of 10

7 The third strand of CoPS work in was outreach. The CoPS team is enthusiastic about having their research used in practical policy situations. CoPS continues to work with U.S. government agencies on a variety of modelling topics: the U.S. International Trade Commission - trade policy, baseline forecasting, model validation; the U.S. Department of Commerce - macro policy, National Export Initiative, biofuels, environmental regulations, illegal immigration; the U.S. Department of Homeland Security - illegal immigration, H1N1 epidemic, terrorism events, suspension of international trade and port closures; the U.S. Department of Energy - greenhouse policies in the electricity sector, renewable fuel standard; and the U.S. Department of Agriculture - biofuels and food prices, low-skilled immigrant agricultural workers. To disseminate its capabilities and results more widely, CoPS conducts training courses throughout the world for potential users of its models. One of these courses takes place each year in Washington DC at the U.S. International Trade Commission. It has been attended by officers from all of the Departments listed above Participants in the course from Homeland Security have been Bryan Roberts, Gary Becker, Marvin Fell and Katie Foreman. REFERENCES Dixon, P.B., B.R. Parmenter, John Sutton and D.P. Vincent, ORANI: A Multisectoral Model of the Australian Economy, Contributions to Economic Analysis 142, North-Holland Publishing Company, 1982, pp. xviii Dixon, P.B., and M.T. Rimmer, Dynamic General Equilibrium Modelling for Forecasting and Policy: a Practical Guide and Documentation of MONASH, Contributions to Economic Analysis 256, North-Holland Publishing Company, 2002, pp. xiv+338. Dixon, P.B. and Maureen T. Rimmer Restriction or Legalization? Measuring the economic benefits of immigration reform, Trade Policy Analysis paper no. 40, Cato Institute, Washington DC, August, 2009, pp. 22. Dixon, P.B. and Maureen T. Rimmer, Validating a detailed, dynamic CGE model of the U.S., Economic Record, 86(Special issue), September, 2010, pp Giesecke, J.A. (2011), Development of a large-scale single U.S. region CGE model using IMPLAN data: A Los Angeles County example with a productivity shock application, Spatial Economic Analysis, Vol. 6(3), pp Giesecke, J.A., W. Burns, A. Barret, E. Bayrak, A. Rose, P. Slovic and M. Suher (2012), Assessment of the regional economic impacts of catastrophic events: CGE analysis of resource loss and behavioral effects of an RDD attack scenario, Risk Analysis, Vol. 32(4), April, pp Verikios, G., Sullivan, M., Stojanovski, P., Giesecke, J. and Woo, G. (2011) The Global Economic Effects of Pandemic Influenza. General Paper No. G-224, Centre of Policy Studies (The Impact Project), Monash University, October. Page 6 of 10

8 2. Publications and Reports CoPS papers in related to its work with CREATE (1) Giesecke, J.A., W. Burns, A. Barret, E. Bayrak, A. Rose, P. Slovic and M. Suher. (2012). Assessment of the regional economic impacts of catastrophic events: CGE analysis of resource loss and behavioral effects of an RDD attack scenario. Risk Analysis, vol 32(4), April, pp Summary We investigate the regional economic consequences of a hypothetical catastrophic event attack via radiological dispersal device (RDD) centered on the downtown Los Angeles area. We distinguish two routes via which such an event might affect regional economic activity: (i) reduction in effective resource supply (the resource loss effect) and (ii) shifts in the perceptions of economic agents (the behavioral effect). The resource loss effect relates to the physical destructiveness of the event, while the behavioral effect relates to changes in fear and risk perception. Both affect the size of the regional economy. RDD detonation causes little capital damage and few casualties, but generates substantial short-run resource loss via business interruption. Changes in fear and risk perception increase the supply cost of resources to the affected region, while simultaneously reducing demand for goods produced in the region. We use results from a nationwide survey, tailored to our RDD scenario, to inform our model values for behavioral effects. Survey results, supplemented by findings from previous research on stigmatized asset values, suggest that in the region affected by the RDD, households may require higher wages, investors may require higher returns, and customers may require price discounts. We show that because behavioral effects may have lingering long-term deleterious impacts on both the supply-cost of resources to a region and willingness to pay for regional output, they can generate changes in regional GDP much greater than those generated by resource loss effects. Implications for policies that have the potential to mitigate these effects are discussed. (2) Verikios, G., M. Sullivan, P. Stojanovski, J. Giesecke, G. Woo (2011) The Global Economic Effects of Pandemic Influenza. General Paper No. G-224, Centre of Policy Studies (The Impact Project), Monash University, October. Summary We analyse the global economic effects of two influenza pandemics that represent extremes along the virulence-infectiousness continuum of possible pandemics: a high virulence-low infectiousness event and a low virulence-high infectiousness event. We do this by applying results from a susceptible infected- recovered epidemiological model to a detailed, quarterly computable general equilibrium model. Our findings indicate that global economic activity will be more strongly affected by a pandemic with high infection rates rather than high virulence rates, all else being equal. At the regional level, regions with a higher degree of Page 7 of 10

9 Researc h Area Referre d Not Referre d PDF Availab le for DHS economic integration with the world economy will be affected more strongly than less integrated regions. (3) Dixon, P., M. Rimmer and B. Roberts, Restricting employment of low-paid immigrants: social welfare implications for legal U.S. wage-earners, Contemporary Economic Policy (forthcoming), Summary Using a detailed CGE model we simulate the effects on the U.S. economy of a tighter-border-security policy that reduces employment of low-paid immigrants. The policy increases wage rates for low-paid legal U.S. workers. On the other hand, it changes the occupational mix of legal employment towards low-paid occupations. We assess the social welfare impact of these labor-market changes using a CES social welfare function. Under what we consider to be a reasonable ethical stance, the positive low-wage effects shown in our CGE simulation are outweighed by the negative occupation-mix effect, leaving the welfare of legal workers reduced by cuts in low-skilled immigration. CREATE PUBLICATIONS BY COPS RESEARCHERS IN 2012 Monash University 1. Giesecke, J.A., W. Burns, A. Barret, E. Bayrak, A. Rose, P. Slovic and M. Suher. (2012). Assessment of the regional economic impacts of catastrophic events: CGE analysis of resource loss and behavioral effects of an RDD attack scenario. Risk Analysis. 2. Verikios, G., M. Sullivan, P. Stojanovski, J. Giesecke, G. Woo (2001) The Global Economic Effects of Pandemic Influenza. General Paper No. G-224, Centre of Policy Studies (The Impact Project), Monash University, October. 3. Dixon, P., M. Rimmer and B. Roberts, Restricting employment of low-paid immigrants: social welfare implications for legal U.S. wage-earners, Contemporary Economic Policy (forthcoming), EA Y Y EA Y Y EA Y Y Page 8 of 10

10 3. Presentations Outreach Wittwer, G., P.B. Dixon and M.T. Rimmer, USAGE-R51 (USA-TERM), a state-level multi-regional CGE model of the U.S. economy, paper presented at the 2012 GTAP Conference, Geneva, June. Summary We set out a quite general specification of the demand side in a bottom-up multiregional CGE model. It quickly becomes apparent that for a detailed model, this general specification raises two problems: (1) coefficient and parameter values need to be estimated at a level beyond what is normally available from statistical sources; and (2) variables with extremely high dimension need to be computed. We show how both these problems are addressed in USAGE-R51 by adapting procedures first devised by Mark Horridge. We set out Horridge s data generating formulas and shows that if data are generated by these formulas and if some other restrictions are imposed to guarantee that Horridge s data assumptions are maintained as the simulated economy moves away from its initial state, then the general system reduces to a much lower dimension system. We argue that Horridge s procedures allow us to manage data and dimensionality problems while not seriously sacrificing realism and policy-relevant detail. The paper describes the database for USAGE-R51and provides illustrative application. J.A. Giesecke and J.R. Madden. (2011) Regional economic impacts in the presence of interregional migration: A CGE explanation. Paper presented to 58th Annual North American Meetings of the Regional Science Association International, Miami FL, November 9-12, [Under preparation for the Journal of Regional Science] Summary We develop and apply a theory of inter-regional gross migration flows that models gradual adjustment of inter-regional wage relativities via changes in gross inter-regional migration flows, within a framework that recognises short-run stickiness in regional wages but long-run regional labour market clearing. We apply the model to a scenario in which there is a rise in the compensating wage premium required by workers in one region. We carefully interpret the results, disentangling the effects resulting from the various mechanisms which link regions. We pay particular attention to the role of inter-regional migration, exploring the implications of a number of migration assumptions popular in the regional CGE literature, and assessing the relative merits of each. We conclude with advice on the desirable properties of a theory describing inter-regional migration in a regional CGE context. J.A. Giesecke, M.W. Griffiths and J.R. Madden. (2012) Development of a large-scale dynamic fiscal CGE model of Florida using IMPLAN data. 43rd Mid-Continent Regional Science Association, 9th Biennial National IMPLAN User s Conference, June 6 June 8, 2012, Bloomington, Minnesota. [Under preparation for Annals of Regional Science] Page 9 of 10

11 Summary In earlier research we developed a comparative static large-scale CGE model of a single U.S. region. In applying this model to the investigation of an RDD terrorism attack, we noted that a dynamic version of such a model would facilitate a better representation of the time path of terrorism-induced behavioural effects. In this paper, we build such a model, developing the comparative static version of our earlier work into a dynamic model. In so doing, we apply the inter-regional migration theory developed in Giesecke and Madden. Page 10 of 10