Computational Methods for Economics

Whether employed by theorists, policymakers or developers, practical applications of the computational models and analytical devices can be used on all levels of economic assessment. The methods themselves are myriad — they range from general algorithms and formulae, to comprehensive state-level modeling, to virtually microscopic sector-based analyses. The results are equally extensive and necessary for truly understanding the systems that build and maintain a macroeconomy.

Keywords Algorithm; Cost Benefit; Empirical; Macroeconomy; Modeling; Sector Analysis

Economics > Computational Methods for Economics

Overview

Samuel Johnson once said that "The use of traveling is to regulate imagination by reality, and instead of thinking how things may be, to see them as they are" (Bartlett, 1919). Indeed, it is true that there are many concepts that, when applied in different analytical environments, act or appear significantly different from their other incarnations. In the case of the Age of Exploration, there were those who viewed the world as flat, ending at the edge of the visible horizon. Of course, for those who set out in ships to the New World (and for humanity as a result of their pioneering), the world became a vastly different place.

In the study of any discipline, there are three methods to employ. The first is the theoretical, one in which hypotheses and conceptual themes are given light. In Johnson's metaphor, theoretical adherents use their imagination to view an environment.

The second is the empirical method, which uses data from experiments, interviews, polls and tests to prove or disprove theories. Empiricists are more rooted in the earth, accepting new ideas only if the evidence of those concepts is truly verifiable.

Bridging the gap between theory and reality is computation. Again using Samuel Johnson's comment, computation is the traveler, inspired by what is theoretically possible and willing to strike out to irrefutably verify or repudiate the idea at hand. Computation rests at the center of the oft-conflicting camps of theory and empiricism. The methods employed using computation result in a new theory, or they may provide evidence that clarifies or discounts previously garnered empirical data. In short, computation's place in an analytical situation is critical and essential for identifying the best possible information.

There are countless forms of methods to employ when assessing trends in economics. Among them are algorithmic formulae, linear modeling and sector-specific modeling. Economists will utilize any of these methods, tailoring them in order to better encompass the topic of study much in the same way that a chemist will add or subtract varying volumes of compounds in order to achieve an experimental result. In this paper, many of the types of computational assessment are, in a general mode, discussed within the broader context of how certain trends and issues are studied.

In economics, the need for both educated theory and empirical data is exceptional. In an ever-changing global economy, many previously established theories and positions have been discounted, and others have yet to be borne. This paper takes a close look at some of the aforementioned computational methods used in the study of economics. Using examples from both the theoretical and empirical arenas of macroeconomics, this author highlights the links built between the two in the practice of economic analysis.

Settling the Debate

Since the latter 20th century and the change into the Millennium, academics, policymakers, business forecasters and observers in general have taken great pains to grasp the world's economic trends. One of the international economic stage's big stars, globalization, has received a particular amount of attention, as links between national economies are becoming more extensive by the day. In some countries, manufacturing has declined as globalization has become the norm. Experts debated the notion that globalization and this trend of "deindustrialization" were linked — that greater free trade, enhanced transportation and communications systems caused disinvestment in national manufacturing industries, which were becoming unnecessary in light of inexpensive imports.

Some studies, hinged on theoretical analyses, pointed to economic development and productivity, not globalization, for the decline in manufacturing in certain systems. However, empiricists, seeing holes in such hypotheses, gave a more careful analysis of this decline, focusing attention on the workforces of each manufacturing industry. By employing a curvilinear model as opposed to one of the more traditional "U-shaped" models, one study revealed a connection; one that is more subtle and therefore more revealing. Globalization, the model determined, causes differentiation between manufacturing sectors, which in turn creates a saturated market. This saturation is the culprit in the decline of manufacturing jobs (Brady, 2006).

The example above provides an illustration of the bridge formed between theory and empirical data. The authors, reviewing a theory that purported a lack of linkage between globalization and manufacturing declination and summarily forming an alternative hypothesis, were able to use the data found in "real-world" systems to verify their position.

Helping to Make Effective Policy

In a world in which linkages are being formed not only among long-standing trade partners and industrialized nations but between so-called "northern" and "southern" states as well, aid for developing nations has become a tremendous component of international diplomacy and relations. Of course, development monies do not come without strings. In fact, any nation that contributes international aid funds to a developing country or government seeks a return on that investment. Hence, tracking the effectiveness of an international development investment is an important part of government policymaking.

On one side of the debate over this issue are those who assert that international development funds do little more than create dependency rather than help generate economic growth. On the opposite end are those who believe that many developing nations do not reach their potential because the funds invested in their growth are insufficient. It is in the analysis of the effectiveness of international aid programs in which an economic method of computation may be useful.

An effective method of analysis in this debate is a sector-specific approach. In other words, economists may focus on the very elements into which international aid is infused: Educational institution-building, financial infrastructure development, environmental protection and other arenas. By studying the growth (or lack thereof) of a sector and the time in which that growth occurs, and factoring in periods of stagnancy (which could be periods in which war, civil unrest or natural disaster occurred), an accurate picture of each sector can become manifest. One study following this methodology applied American aid, in varying forms (such as conditional grants and unrestricted aid) to this sector analysis. The authors' results paint an extremely interesting illustration of the most effective forms of international aid for nations of varying size, economic status and geo-political status (Dovern, 2007).

Addressing Impacts

One of the timeliest of issues facing municipalities in the United States is whether or not to embrace casinos and gaming. In recent years, the number of states to allow casinos and/or casino-style gambling has jumped to 40, with several others presently considering following suit (Cauchon, 2007). The potential economic benefits that states see are sizable — millions of dollars in revenues that could potentially serve as a boon for economically distressed cities and regions. Here too, computational methods may be employed to study not only the monetary potentials (or losses), but the social consequences as well, of legalized gambling.

Using a sector analysis approach, one study reviewed several economic factors. One of the most significant elements studied was that of unemployment in the state of Michigan. Popular opinion among advocates maintains that legalized casino gambling will bring jobs to depressed areas. However, a computational study of the empirical data in this arena paints a rather different picture.

In fact, the study suggests, the introduction of casino gambling has been relatively innocuous in mitigating high unemployment rates in certain sections of that state. Unemployment rates remained static in times of economic stagnancy, and moved downward in boom times, in both areas with casinos and without (Koo, 2007).

This analytical work serves as an illustration of the usefulness of a sector analysis. In this case, a popular theory (that legalized gambling generates much-needed jobs in areas experiencing poor fiscal health) that serves as a rallying cry for advocates is proven invalid in the areas under study.

Studying Consumer Behavior

The ever-changing global economy logically creates questions among economists as to how short- and long-term trends affect consumer behavior. In one arena, the increase in liberalized economies has created diversified industrial bases as well as higher incomes. Still, if incomes are higher, how does this change affect demand and consumption? In a similar vein, with greater wealth across demographic strata, what impact does this trend have on price?

One study, offered by Ray Barrell and Philip Davis, analyzes the impact of economic liberalization on consumption among the US, Germany, the UK, France, Japan, Canada and Sweden. Using empirical data models that center on changes to wealth and income, credit restraints and other reductions in liquidity, interest rates and, of course, pricing, the authors find that economic liberalization, and in particular, the removal of fiscal restraints, does have a significant impact on consumption. In fact, the study shows, in a general sense, that removal of those liquidity constraints helps bolster individual wealth and reduces elasticity (Barrell, 2007).

In this case, the methodology used depended not on preestablished theoretical frameworks, as such models disregarded certain elements of an economy that can be a factor in economic development. By focusing on a wide range of components (such as interest rates) and providing individual models of similar but separate economic systems (for example, not combining data from American and Swedish systemic conditions), the authors were able to locate certain truths about the relationships between liberalization and consumer behavior.

Finding Chaos in the Order

The study described above raises another interesting arena in which economic computational methods can prove most useful. In the previous examples, I have demonstrated how computational methods for economics can be used to prove or disprove certain theoretical conclusions. In the following example, however, it becomes clear that theory is slower to appreciate certain trends in economics than is empirical application. Thus, these computational methods can be employed in reverse, bringing empirical data back to the theoretical side of macroeconomics and helping spur the development of new, more updated hypotheses.

In the course of studying the nexus between supply and demand, pricing, there are a number of variables that exist which may play "x-factor" in the establishment of equilibrium in a particular system. In fact, not every market operates exactly the same in a "real world" setting. At the theoretical level, market behavior is often analyzed under a similar set of "rules," parameters and formulae. When one takes into account empirical data, however, certain external and previously unanticipated elements can cloud the "clean" model found in a theoretical framework, much in the way atmospheric conditions can impact an experiment once it is removed from the laboratory.

One analysis of monetary policy takes issue with an oversimplified view by theoretical economists that the focal point of fiscal policy should be the relationship between supply and demand and, at the center of that relationship, the establishment of equilibrium. However, there are variables that play a significant role in this general relationship and which may cause unanticipated anomalies within certain models. Among these variables are such factors as floating exchange rates and stock returns. The authors correctly assess that, in light of the inconsistencies that arise when utilizing conventional modeling to analyze certain free markets, a broader and more "outside the box" approach is necessary. Such an action can help foster better modeling at an academic level and, as a result, provide better data-collection methodologies:

Such approaches may concurrently aid modeling as well as establishing effective monetary policy.

Cost Benefit Analyses

When a development project, such as the construction of a roadway, rehabilitation of an existing building or even the construction of an industrial site, is being considered, it is critical to conduct another form of computational economic analysis. There are a number of methods that can be utilized, including economic impact assessments, life cycle cost analysis and a cost-benefit analysis. Each weighs the potential benefits against the risks involved in a major undertaking such as a large-scale project.

The first of these options is the economic impact assessment (EIA), which entails studying the effect of the project on surrounding infrastructure, systems and people. The second, a life cycle cost analysis, examines the expected costs of the project over the entire life of the project. The third, the benefit-cost analysis, examines the costs associated with the investment in direct comparison to the long-term expenses as well as the goal of the project itself (Gabler, 2004).

Any major construction project raises eyebrows. However, the construction of a sports stadium tends to cause even more public scrutiny, as taxpayers are usually wary of using public funds to finance a private development undertaking. In northern New Jersey, a $750 million redevelopment project proposal involving a new football stadium complex to replace the 30 year-old Giants Stadium has been approved. The road to that approval in 2005 was no easy one — an extensive, 35-page project summary was drafted and included a complex EIA. Of course, the cost-benefit analysis has yet to be formally conducted, which is much to the consternation of onlookers. One report suggested that, in the long-term, the loss of revenue to sporting events and concerts at the complex far outweigh any tax revenues (Mansnerus, 2005). Absent a thorough benefit-cost analysis and life cycle analysis, both of which have yet to be introduced by an unbiased source, the controversy surrounding this project will likely continue until all questions are answered satisfactorily.

Conclusion

Between the realm of the theoretical study of economics and the "real world" trends in an economic system are the methods used to study each. Computational methods for economic analysis help provide clarity and even veracity of academic hypotheses. Likewise, such methods can help economists formulate models by which to study empirical data.

In many situations, use of any of the myriad of linear, nonlinear, mathematical or sector-specific models in computational analysis can help settle an issue that rests at the very core of a major economic trend. As is the case with the impact of globalization on demand and, ultimately, price equilibrium, empirical analysis of the individual factors that comprise consumer behavior may fill in many of the holes that exist in the ever-evolving study of a global economy.

The computational methods employed to study economic systems can also be used to examine the effectiveness of a national investment program. As shown in this paper, such methodologies can be used to gauge the effectiveness of international aid and development investments. Such data can be critical for establishing long-lasting and mutually beneficial interstate relationships.

Furthermore, economic analysis employing these techniques may help provide evidence of critical trends within an issue area. As shown in the case of the issue of legalized gambling and casinos, the assertion by some that gaming will create jobs is not entirely valid when one takes into account many of the locales in which such gaming already takes place. While this evidence will likely not satisfy either conflicting party, the information it reveals is nonetheless critical and, most importantly, well-founded.

With the global environment establishing interstate linkages in ways and degrees never before seen in human history, the discipline of studying these systems is still quite new. The models that rest between the theoretical and empirical realms are therefore very useful to any researcher. In the case shown in this paper, the Keynesian approach of seeking to understand consumer behavior may benefit from the employment of the very computational methods discussed in this essay.

Computational methodologies for studying economics can also be useful in opening doors for theorists. As this essay has demonstrated, previously unanticipated factors and trends, uncovered with the study of empirical data, can be useful in developing new conceptual models. Hence, the flow of information between the theoretical and empirical arenas goes in both directions with the use of such computational methods. However, the world of analysis is not limited to the traditional approaches, particularly when new and exogenous empirical data (and the methods by which it is tracked) becomes manifest.

Finally, this paper shows the use of economic computational methods not only for the purposes of national and international policies, but for the use of local activities as well. In the case of major construction projects such as the development of the Meadowlands in New Jersey, a series of evaluations of the project's impacts on the region and its residents was not only important for the project's progress — it was required by local, state and federal law.

Whether employed by theorists, policymakers or developers, practical applications of the computational models and analytical devices can be used on all levels of economic assessment. The methods themselves are myriad — they range from general algorithms and formulae, to comprehensive state-level modeling, to virtually microscopic sector-based analyses. The results are equally extensive and necessary for truly understanding the systems that build and maintain a macroeconomy.

Terms & Concepts

Algorithm: Computational method used to systematically solve a given problem.

Cost Benefit: Study that weighs expected expenses versus potential positive returns.

Empirical: Based on data obtained in the field, not in theory.

Macroeconomy: An economic system that includes supply and demand elements.

Modeling: Research method in which trends and subjects are connected for purposes of empirical study.

Sector Analysis: Methodology in which elements of a macroeconomy are separated and studied individually.

Bibliography

Barrell, R. & Davis, P. (2007). Financial liberalisation, consumption and wealth effects in seven OECD countries. Scottish Journal of Political Economy, 54, 254-267. Retrieved October 31, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=24669736&site=bsi-live

Bartlett, J. (1919). Familiar quotations. Bartleby.com. Retrieved October 29, 2007, from http://www.bartleby.com/br/100.html.

Brady, D. & Denniston, R. (2006). Economic globalization, industrialization and deindustrialization in affluent democracies. Social Forces, 85, 297-329. Retrieved October 30, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=22490744&site=bsi-live

Cauchon, D. (2007, August 13). Cities gamble on casinos for tax revenue. USA Today (Online Edition). Retrieved October 30, 2007, from http://www.usatoday.com/news/nation/2007-08-13-casinos%5fN.htm.

Chen, S., & Wang, S. G. (2011). Emergent complexity in agent-based computational economics. Journal of Economic Surveys, 25, 527-546. Retrieved November 24, 2013, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=60602327&site=ehost-live

Cuthbertson, K. & Nietzsche, D. (2007). Monetary policy and behavioural finance. Journal of Economic Surveys, 21, 935-969. Retrieved October 31, 2007 from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=27013812&site=bsi-live

Dovern, J. & Nunnenkamp, P. (2007). Aid and growth accelerations: An alternative approach to assessing the effectiveness of aid. Kyklos, 60, 359-383. Retrieved October 30, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=25779921&site=bsi-live

Gabler, E. (2004). Economics of investment. Roads and Bridges, 42, 66. Retrieved October 31, 2007, from EBSCO Online Database Academic Search Complete. http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=12657046&site=ehost-live

Koo, J., Rosentraub, M. S., & Horn, A. Rolling the dice? Casinos, tax revenues, and the social costs of gaming. Journal of Urban Affairs, 29, 367-381. Retrieved October 30, 2007, from EBSCO Online Database Academic Search Complete. http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=26771764&site=ehost-live

Lewis, S. C., Zamith, R., & Hermida, A. (2013). Content analysis in an era of big data: A hybrid approach to computational and manual methods. Journal of Broadcasting & Electronic Media, 57, 34-52. Retrieved November 24, 2013, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=86010175&site=ehost-live

Mansnerus, L. (2005, April 23). New Meadowlands stadium is approved for the Giants. New York Times (Online Edition). Retrieved October 31, 2007, from http://www.nytimes.com/2005/04/23/nyregion/23giants.html.

Warner, A. G., & Caliskan-Demirag, O. (2011). An agent-based computational economics approach to technology adoption timing and the emergence of dominant designs. Journal of Business & Economics Research, 9, 107-119. Retrieved November 24, 2013, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=59159517&site=ehost-live

Suggested Reading

Camilleri, D., Mollicon, P. & Gray, T.G.F. (2007). Computational methods and experimental validation of welding distortion models. Proceedings of the Institution of Mechanical Engineers — Part L — Journal of Materials: Design & Applications, 221, 235-249. Retrieved November 2, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=27019107&site=ehost-live

Dillard, J. & Nissen, M. (2007). Computational modeling of project organizations under stress. Project Management Journal, 38, 5-20. Retrieved November 2, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=24844782&site=ehost-live

Missaglia, M. (2006). Dynamic general equilibrium modeling. Computational methods and applications. Journal of Economics, 88, 207-209. Retrieved November 2, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=21887748&site=ehost-live

Poe, G.L., Giraud, K.L. & Loomis, J.B. (2005). Computational methods for measuring the difference of empirical distributions. American Journal of Agricultural Economics, 87, 353-365. Retrieved November 2, 2007, from EBSCO Online Database Business Source Complete. http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=16702013&site=ehost-live