Economic question

The Independent Institute, 100 Swan Way, Oakland, CA 94621 • 800-927-8733 • Fax: 510-568-6040

Card No.

Name

Organization

Street Address

City/State/Zip/Country

Signature Email

Exp. Date CVC Code

Telephone No.

Title

prOmO CODE IrA1402

SUBSCRIBE NOW and Get CRISIS AND LEVIATHAN FREE!

Subscribe to The Independent Review and receive your FREE copy of the 25th Anniversary Edition of Crisis and Leviathan: Critical Episodes in the Growth of American Government, by Founding Editor Robert Higgs. The Independent Review is the acclaimed, interdisciplinary journal by the Independent Institute, devoted to the study of political economy and the critical analysis of government policy.

Provocative, lucid, and engaging, The Inde- pendent Review’s thoroughly researched and peer-reviewed articles cover timely issues in economics, law, history, political science, philosophy, sociology and related fields.

Undaunted and uncompromising, The Independent Review is the journal that is pioneering future debate!

Student? Educator? Journalist? Business or civic leader? Engaged citizen? This journal is for YOU!

SEE mORE at: INdEpENdENt.ORG/tIROFFER

SUBSCRIBE to The Independent Review NOW and Receive a FREE copy of Crisis and Leviathan

OR choose one of the following books: q

q YES! Please enroll me with a subscription to The Independent Review for: q Individual Subscription: $28.95 / 1-year (4 issues) q Institutional Subscription: $84.95 / 1-year (4 issues)

q Check (via U.S. bank) enclosed, payable to The Independent Institute q VISA q American Express q MasterCard q Discover

q

The Challenge of Liberty Classical Liberalism Today

Edited by Robert Higgs and Carl Close

q

Lessons from the Poor Triumph of the

Entrepreneurial Spirit Edited by Alvaro Vargas Llosa

q

Living Economics Yesterday, Today

and Tomorrow By Peter J. Boettke

q

The Terrible 10 A Century of

Economic Folly By Burton A. Abrams

Government behind the Wheel

More a Matter of Politics Than of Economics

F

JIM F. COUCH, PHILIP A. BURTON, KEITH D. MALONE, AND

DAVID L. BLACK

Closed [car] dealerships across the country deserve a transparent review

of their termination and the right to get back in business if they were

terminated on faulty grounds.

—Senator Richard Durbin (D–Ill.)

(qtd. in “Congress Plans Remedy” 2009)

My Administration is committed to creating an unprecedented level of

openness in Government. We will work together to ensure the public

trust and establish a system of transparency, public participation, and

collaboration. Openness will strengthen our democracy and promote

efficiency and effectiveness in Government.

—Barack Obama (White House Press Office 2009)

Jim F. Couch is a professor of economics; Philip A. Burton is an assistant professor of quantitative methods; Keith D. Malone is an assistant professor of economics; and David L. Black is an assistant professor of economics. All are at the University of North Alabama.

The Independent Review, v. 15, n. 4, Spring 2011, ISSN 1086–1653, Copyright © 2011, pp. 577–591.

577

T he business of America is business,” President Calvin Coolidge once

famously opined, and no industry better represented American prowess than

automobile manufacturing.1 But that prowess has slowly eroded to the point

that more than half of the cars Americans buy are “foreign” (“America’s Other Auto

Industry” 2008).2 Domestic manufacturers—the so-called big three—have lost

market share and face perhaps insurmountable financial troubles.

The big three’s problems have several main sources, among others: high

labor costs (for every United Auto Worker member working at a car manufac-

turer, three collect generous retirement benefits [“America’s Other Auto Indus-

try” 2008]); poor quality, whether perceived or real; inefficiency (in 1995, for

example, a General Motors [GM] automobile required forty-six hours to build,

whereas a Toyota required 29.4 [“America’s Other Auto Industry” 2008]); and

relatively low resale value. Escalating fuel costs worked against the sale of trucks

and sport utility vehicles—something domestic producers did well and had hung

their hats on. The financial crisis further exacerbated the problems Chrysler and

GM faced to the point that both faced certain bankruptcy as 2008 came to a

close. Domestic firms went from being the very symbol of American manufactur-

ing strength to financial ruin.

Business has changed since Coolidge’s day as well, with government stepping

in to bail out struggling firms as never before. President George W. Bush,

encouraged by President-elect Barack Obama, provided funds for GM and

Chrysler from the Troubled Assets Relief Program (TARP) at the end of his

term—a fund authorized by Congress to bail out banks and financial institutions.

Bush defended his action by saying: “In the midst of a financial crisis and a

recession, allowing the U.S. auto industry to collapse is not a responsible course

of action” (qtd. in Isadore 2008).3 Bush was simply keeping the firms afloat so

that his successor could find a more permanent solution for the moribund

manufacturers.

President Obama took responsibility for restoring the industry. In March

2009, he asserted: “Year after year, decade after decade, we’ve seen problems

papered over and tough choices kicked down the road, even as foreign competi-

tors outpaced us. Well, we’ve reached the end of that road! And we, as a nation,

cannot afford to shirk responsibility any longer. Now is the time to confront our

1. The statement is from a speech by Coolidge called “The Press under a Free Government,” delivered to the American Society of Newspaper Editors in Washington, D.C., on January 17, 1925. Coolidge appar- ently actually stated: “After all, the chief business of the American people is business.” The entire speech indicates that he was not as materialistic as his detractors have asserted. See http://www.calvin-coolidge. org/html/the_business_of_america_is_bus.html.

2. International manufacturers employ more than one hundred thousand U.S. workers building Hondas, BMWs, Mercedes, Toyotas, Hyundais, and other foreign brands.

3. Owing in part to the timeliness of the issue and in part to the subject matter, we rely to a greater extent than is typical on information from the popular press. The reader should be aware that these accounts are not subject to the same scrutiny as refereed sources.

578 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

problems head-on and do what’s necessary to solve them” (White House Press

Office 2009).4

The president formed an automobile task force cochaired by Treasury secretary

Tim Geithner and National Economic Council director Larry Summers, and Chrysler

and GM were instructed to submit restructuring plans to the group for approval. Task

force members included “the secretaries of Transportation, Commerce, Labor,

Energy, the director of the Office of Management and Budget, the administrator of

the Environmental Protection Agency, the director of the White House Office of

Energy and Climate Change and the chair of the Council of Economic Advisors”

(Shepardson and Trowbridge 2009).

Steven Rattner, an investment banker, was selected to head the team.

Rattner acknowledged the complexity of the task and his team’s inexperience in

regard to the automobile industry, confessing, “We’ve learned a lot about how car

dealers work, and how companies get paid when they sell a car to a dealer, and why

there are a certain number of dealers more than are optimal” (qtd. in King and

Stoll 2009).

The president seemed pleased with the automobile task force’s progress. In a

radio interview with Michael Smerconish, Obama criticized Bush’s actions and

praised his own efforts. “The only thing that we did was rather than just write GM

and Chrysler a blank check, we said, ‘You know what, if you’re going to get any more

taxpayer money, you’ve got to be accountable’” (“Obama Is Interviewed” 2009).

The task force rejected Chrysler’s initial viability plan but later accepted its second

plan. The company filed under Chapter 11 of the U.S. Bankruptcy Code and as a

condition of the bankruptcy sought to close 789 of its 3,181 authorized dealerships.

Dealerships received a list of all the firms slated for closure and a message that stated:

“With regret, this letter is to inform you that on May 14, 2009, we are filing a motion

in bankruptcy court rejecting the Sales and Service Agreement (s) between Chrysler

Motors LLC and the dealerships listed above” (qtd. in Valdez-Dapena 2009).

Chrysler officials noted that the decision to close dealerships—in particular

which franchises would be terminated—was difficult but necessary. James E. Press,

vice chairman of Chrysler, stated, “This has been the most difficult business decision I

have ever personally had to take. But the decision had to be made. They [sic] were

gut-wrenching, but absolutely necessary for Chrysler’s survival” (qtd. in Becker

4. One of the major philosophical debates on the federal government’s “acquisition” of domestic automo- bile manufacturers pertains to the fundamental role of government. Although most elected federal officials avoided this question during the weeks leading up to the decision to acquire these companies, the question looms large. The larger question remains as to whether the government can or should acquire the assets of a failing major company, but the subtext of the “takeover” was not foreseen—the federal government now has conflicting, dual roles as both a competitor and a referee. An appropriate analogy would be to an umpire of a baseball game who wears the emblem of the team he prefers to win. This dual role reduces the trust and transparency that free markets require. Under this new “market” arrangement, all federal gov- ernment regulations are suspect. The recent congressional hearings about the mechanical problems various Toyota models have experienced indicate that a third-way blend of government and corporations establishing a cooperative relationship is ill suited for free markets.

GOVERNMENT BEHIND THE WHEEL F 579

VOLUME 15, NUMBER 4, SPRING 2011

2009). The exact criteria used to determine which Chrysler dealerships got the ax

were not clear. Chrysler’s executive vice president for North American sales, Steven

Landry, asserted that the process was a function of the numbers: “The decision,

though difficult, was based on a data-driven matrix that assessed a number of key

metrics” (qtd. in Valdez-Dapena 2009).

Various observers suggested that dealership profitability, customer-satisfaction

surveys, whether the full line of Chrysler products was offered to shoppers, and

dealership saturation in a particular area contributed to the decision to close a dealer-

ship. Heritage Foundation researchers argued that reducing intrabrand competition

should be a significant factor in the decision-making process: “A surplus of franchises

means dealers for the same manufacturer end up competing among themselves,

resulting in lower returns across the board” (Sherk and Gattuso 2009, 2).

Nevertheless, many dealers, facing the loss of their franchise, complained bitterly

that their firms were profitable and that “the criteria being used to determine which

dealerships survive is [sic] not clear” (qtd. in Tapscott 2009). Others suggested that

politics influenced the decision.

This assertion gained plausibility when it transpired that all of the dealerships in

the McLarty-Landers-Johnson chain escaped closure. Robert Johnson, one of the

firm’s owners, is the founder of Black Entertainment Television and is a heavy con-

tributor to the Democratic Party. “Mack” McLarty, another owner, was an aide to

former president Bill Clinton. Accusations spread that dealerships owned by individ-

uals with ties to the Republican Party were being targeted for closure.

In addition, adding fuel to the fire, people whose chief concern was politics

rather than economics provided assistance to the task force. “Senior aides advising

the task force include . . . former Obama campaign aides who are already hard at work

reviewing a number of issues related to the restructuring of General Motors Corp.

and Chrysler LLC” (Shepardson and Trowbridge 2009).

The Treasury Department responded with a statement claiming that the

government had played no role in selecting the dealerships whose franchise agreement

would be terminated. The “[d]epartment had no role in choosing which contracts

would be dropped or in the number dropped” (qtd. in Valdez-Dapena 2009).

Nevertheless, attorney Leonard Bellavia, who represents a number of closed

dealerships and has deposed senior Chrysler officials, told Reuters, “It became clear

to us that Chrysler does not see the wisdom of terminating 25 percent of its dealers. It

really wasn’t Chrysler’s decision. They [Chrysler officials] are under enormous pres-

sure from the President’s automotive task force” (qtd. in Tapscott 2009).5

5. At least one Chrysler executive has changed his tune: “[A]s Chrysler continues to report dismal sales, industry experts and even a former executive question whether the company blundered by dropping so many dealers and giving them just over three weeks to wind down operations. Jim Press, who served as Chrysler’s president and vice chairman until last month, said in an interview that he personally fought the dealership closings. ‘I saw it fraught with terrible issues and short-term sales cost as well as dislocation of customers,’ Mr. Press said. ‘Dealers are [Chrysler’s] only customers, the reason we are in business. How do you eliminate your customer?’” (Maltby 2009).

580 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

Political Considerations

The Heritage Foundation, intrigued by the notion that political contributions might

have shaped the decision to terminate Chrysler dealerships, conducted a study titled

Closing Car Dealerships: A Matter of Economics, Not Politics (Sherk and Gattuso

2009). The authors analyzed data provided in Chrysler’s Chapter 11 bankruptcy

filings with the U.S. Bankruptcy Court for the Southern District of New York. These

filings included the names of all dealerships—both those that were to remain open

and those slated for closure—as well as each firm’s address and owner(s).

The study’s authors compared the political contributions of the majority owner

of each dealership to the list of closures to determine if “Republican-leaning dealers

were treated more harshly than those that supported Democrats.” They concluded:

“[C]laims that the method by which dealers were selected was biased appear to be

unfounded, with no correlation between political contributions and terminations”

(Sherk and Gattuso 2009, 5).

A Matter of Politics, Not Economics

Public-choice theory asserts that politicians are motivated by a desire for reelection

and that this desire shapes which policies are pursued and how they are implemented.

Thus, the public’s outcry that government actions seldom solve real problems is

misguided because the policy goal is not to provide a solution, but instead to build a

winning coalition for the next election. In addition, government agencies—Obama’s

automobile task force certainly qualifies—are not immune to political influence. Both

the executive branch and the legislative branch of government use the apparatus of

government for reelection purposes. In short, politics trumps economics. A vast

literature supports this assertion.

Several scholars have found that politics played an important role in the alloca-

tion of New Deal appropriations during the Great Depression (Wright 1974; Ander-

son and Tollison 1991; Couch and Shughart 1998). Despite Roosevelt’s eloquent

speeches about helping those in distress, economic need was virtually ignored when

dollars were distributed across the nation. Instead, empirical evidence supports the

claim that there was a reelection strategy to the allocation.

Economist William Shughart adds, “One key conclusion of public choice is that

changing the identities of the people who hold public office will not produce major

changes in policy outcomes” (Shughart 2008). Thus, the results of a Mercatus

Center analysis of the Obama administration’s distribution of stimulus funds should

not be surprising. Examining fourth-quarter expenditures from the American Recov-

ery and Reinvestment Act of 2009—Obama’s stimulus package totaling more than

$800 billion in expenditures and, to a much lesser extent, tax cuts—Jerry Brito and

Veronique de Rugy found “no correlation between economic indicators and stimulus

funding. Preliminary results find no effect of unemployment, median income, or even

GOVERNMENT BEHIND THE WHEEL F 581

VOLUME 15, NUMBER 4, SPRING 2011

mean income on stimulus funds allocation” (2009, 6). Instead, the authors deter-

mined that stimulus dollars flowed disproportionately to Democratic-controlled con-

gressional districts. Like FDR, the current administration has put political

considerations ahead of economic need.

Bureaus and agencies succumb to political pressure as well. Jim Couch and

associates (1999) found that the Internal Revenue Service (IRS) conducted fewer

audits in states that had delivered more votes to Clinton. Marilyn Young and

associates (2001) also found a political component to IRS enforcement activity.

Thomas Garrett and Russell Sobel (2003) note that Federal Emergency Manage-

ment Agency disaster expenditures have been guided at least in part by the current

president’s reelection strategy. Enforcement by the Environmental Protection

Agency (Couch, Williams, and Wells 2008) and by antitrust authorities (Faith,

Leavens, and Tollison [1982] 1995) has likewise been found to be politically

motivated. Even the allocation of troops during the Vietnam War was discovered

to be influenced by politics: more troops from states with less political clout were

deployed to more dangerous locations and in more risky activities (Goff and

Tollison 1987).

Empirical Evidence: Dealership Closure

Dealerships provide both direct and indirect employment opportunities; advertising

revenue for local media; and generous tax revenue for city, county, and state govern-

ments derived from the sale of big-ticket items. Moreover, they are frequently called

on to meet community philanthropic needs. Closure of a dealership, whether justified

or not, is obviously unpopular with residents in the surrounding area. A politician

seeking reelection would have a strong incentive to see that his supporters are not

subjected to the negative effects of a closure.

Obama administration officials were quick to point out that they did not have a

voice in the determination of which dealerships would be closed. They claimed that

Chrysler officials made the call. Both the administration and Chrysler executives

asserted that economics drove the calculation. An analysis of the data provides evi-

dence as to whether the administration sought to shield its supporters from the

difficulties associated with the closure of dealerships.

Data Analysis

We began by determining the percentage of dealerships in each state that were closed.

The percentage varied widely from state to state, ranging from a high of 40 percent of

total dealerships closed in North Dakota, Utah, and West Virginia to a low of only 10

percent of total dealerships closed in Vermont (Alaska actually had no dealerships

closed). These data appear in table 1.

582 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

Table 1 Percentage of Dealerships Closed, by State

Percentage of

Dealerships

Closed

Percentage of

Dealerships

Closed Full Line 1, 2*

Percentage of

Dealerships

Closed Full Line 3, 4*

Alabama 31.6 70.0 17.9

Alaska 0.0 0.0 0.0

Arizona 14.7 10.0 16.7

Arkansas 22.9 100.0 15.6

California 26.0 39.2 15.5

Colorado 27.5 29.4 26.1

Connecticut 16.3 35.3 3.8

Delaware 30.0 60.0 0.0

Florida 34.3 63.6 10.9

Georgia 17.8 50.0 11.5

Hawaii 14.3 0.0 14.3

Idaho 13.6 50.0 5.6

Illinois 27.7 59.3 9.0

Indiana 22.5 56.0 9.4

Iowa 25.0 53.8 19.7

Kansas 27.5 75.0 18.6

Kentucky 14.5 50.0 10.2

Louisiana 39.5 50.0 38.5

Maine 20.0 100.0 0.0

Maryland 34.7 61.5 4.3

Massachusetts 25.5 39.1 12.5

Michigan 26.0 57.6 6.3

Minnesota 25.0 33.3 22.4

Mississippi 15.6 50.0 10.7

Missouri 30.5 69.6 15.3

Montana 16.7 100.0 6.3

Nebraska 23.5 100.0 16.1

Nevada 35.7 50.0 20.0

New Hampshire 25.0 66.7 7.1

New Jersey 35.8 49.1 7.7

New Mexico 13.6 40.0 5.9

New York 17.7 52.6 5.5

North Carolina 18.1 35.3 12.7

(continued)

GOVERNMENT BEHIND THE WHEEL F 583

VOLUME 15, NUMBER 4, SPRING 2011

As noted previously, various economic factors and dealership characteristics have

been cited as possible reasons for closing certain Chrysler dealerships. We combined

both individual dealership characteristics and political factors in our analysis to deter-

mine which of these variables played a role in the calculation of franchise termination.6

Table 1 (Continued)

Percentage of

Dealerships

Closed

Percentage of

Dealerships

Closed Full Line 1, 2*

Percentage of

Dealerships

Closed Full Line 3, 4*

North Dakota 40.0 85.7 15.4

Ohio 31.3 63.3 15.3

Oklahoma 21.3 42.9 17.5

Oregon 14.3 22.2 11.1

Pennsylvania 30.1 53.8 8.2

Rhode Island 33.3 100.0 0.0

South Carolina 23.9 40.0 16.1

South Dakota 29.2 55.6 13.3

Tennessee 25.5 66.7 17.4

Texas 24.9 52.6 13.6

Utah 40.0 100.0 28.6

Vermont 10.0 25.0 0.0

Virginia 37.7 61.3 18.4

Washington 31.1 54.5 8.7

West Virginia 40.0 62.5 25.0

Wisconsin 16.0 63.2 4.9

Wyoming 26.3 66.7 18.8

*The “full-line” variable refers to the number of lines of Chrysler products a dealership offers.

A value of 1 indicates the dealership offers only a single line of Chrysler products; a value of

4 indicates that the dealership offers the complete line of Chrysler products (three plus trucks).

6. The selection of explanatory variables in the analysis was guided by specific factors that Chrysler officials mentioned and by political variables standard in the literature. Thus, we examine intrabrand competition, the number of products offered by each dealership, and a proxy for dealership profitability—namely, the unemployment rate. At first glance, interbrand competition might have contributed to the decision to close a Chrysler franchise. However, the literature suggests that competing dealerships cluster together. “Dealer- ships have become concentrated into a small number of large suburban clusters with the spacing between clusters influenced by territorial security concerns for dealers selling the same make of automobile. The strong clustering tendency of dealerships has been attributed to comparison shopping opportunities for consumers and advertising advantages to sellers” (Lord 1992, 283). The term auto mile has been used to describe a collection of competing automobile dealerships that cluster together to market their product better. Hence, empirical evidence suggests that firms seek out their competitors and locate in close proximity to them rather than avoid them.

584 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

We based our analysis on public-choice theory. Thus, rather than suggesting that

political contributions by the owner of each closed dealership to the political party

opposing the current administration were paramount in the closure decision, as the

Heritage study does,7 we employed a more direct measure, reelection efforts, which is

standard in public-choice literature when analyzingthe influence ofthe executive branch.

The percentage of the vote cast for Obama in each state served as a political

variable in our investigation (for example, we entered the same Obama vote percent-

age for each dealership in Alabama). Electoral votes per capita at the state level and

party of the congressperson in the U.S. House of Representatives in each dealership’s

district are included as additional measures of political influence.

We also factored individual dealership characteristics into our analysis. Chrysler

officials noted that the decision to terminate a dealership franchise was based on a

number of factors, but they specifically mentioned dealerships’ profitability and level

of customer service. These variables unfortunately are proprietary in nature, so our

investigation admittedly suffers from their exclusion. We made an effort to account

for profitability by including unemployment as a variable. Thus, the unemployment

rate functions as a proxy for profitability in the analysis. If profitability is indeed a

pertinent issue, then dealerships located in economically depressed areas—those with

high unemployment rates—were more likely to be closed.

Additional dealership characteristics included in the study were the number of

Chrysler products that each dealership offered for sale and dealership saturation in a

particular area. It was expected that dealerships with more extensive offerings and

larger customer bases were more likely to escape closure.

Our first model examined the percentage of total dealerships (data from table 1)

in each state that was closed, so we investigated data at the state level first. Next we

focused the analysis on individual dealerships. We analyzed the data to determine

which factors played a statistically significant role in the decision to terminate a

dealership franchise. A more technical analysis of the data and a description of the

statistical models appear in the appendix to this article.

Both the state-level analysis and the individual-dealership analysis suggest that a

relationship did exist between not offering a full line of Chrysler products and the

likelihood of losing a franchise. Chrysler spokespersons indicated before the closures

that this factor would be a determining one, and we found evidence that it was indeed

considered in their calculation.

We found little evidence, however, to support the notion that Chrysler officials

sought to limit intrabrand competition. In fact, the results from the dealership-level

analysis indicate that Chrysler franchises facing less competition in the county were

more likely to be closed, whereas dealerships in more heavily saturated areas were

more likely to remain open.

7. The Heritage study examines whether a dealership remained open or was closed and the dollar amount do- nated to candidates, but it does not control for other factors associated with the decision to terminate franchises.

GOVERNMENT BEHIND THE WHEEL F 585

VOLUME 15, NUMBER 4, SPRING 2011

Based on the empirical analysis, we also found that the economic health of the

area surrounding the dealership did not affect the decision to terminate a dealership.

Our proxy for economic vitality in the county, the unemployment rate, was not

significantly related to dealership closure.

Thus, of all the factors that various officials offered to explain how dealerships to

be closed were selected, only “offering the full line of Chrysler products” was valid.

None of the other explanations is supported by the analysis. However, we did find

evidence that politics played a role.

After controlling for other factors thought to influence the decision to close, we

found evidence that dealerships were more likely to remain open in states that offered

political support for Obama.8 This result is consistent in both the state-level and the

dealership-level analyses. The evidence therefore suggests that the Obama adminis-

tration was concerned with more than Chrysler’s long-term viability—namely, it was

concerned with protecting Obama’s supporters from the negative consequences of a

dealership closure.

Conclusion

Our results strongly support the notion that the Obama administration—either

directly through the automotive task force or indirectly by putting pressure on

Chrysler executives—played a key role in the selection of which dealerships would be

closed. Although some observers argue that the government can effectively regulate

the private sector, our results suggest that the temptation to succumb to political

considerations is strong and, indeed, cannot be avoided.

All of the model specifications we estimated support a political motivation behind

the selection of dealerships slated for closure.9 President Obama ran on a platform of

change and promised a more open, efficient, and effective government. In the case of

8. A more innocuous explanation can account for the relationship between Obama and dealership closure. Suppose both Obama support in a state and dealership closures in a state are correlated with income. States with higher incomes may have given Obama greater political support and provided a more profitable environment for dealerships. Thus, the significance of the Obama variable may not indicate a political motivation but rather simply reveal that the Obama variable is a proxy for high income. We also tested for this possibility, and the evidence supporting the political hypothesis remains strong. Although state median income is positively and significantly related to state support for Obama, the percentage of dealerships closed in each state and median income are unrelated. The technical detail of the relationship among these variables is provided in the appendix.

9. Public choice assumes that decisions are influenced by the desire to build a strong reelection coalition. However, this behavior can manifest itself in several forms: politicians may choose to reward their loyal supporters or may instead attempt to influence swing voters. We entered the percentage of the vote for Obama in the model tests to see if supporters were favored. We also entered a second variable in the model to capture efforts to win swing voters. In order to test this hypothesis, we calculated a measure of electoral importance for each state. The variable was calculated as Ei ¼ EViP� 1 � 4� Pi � 0:5ð Þ2

� � , where EVi is the

number of electoral votes per state and Pi is the percentage margin of victory for President Obama per state. Thus, a state in which the voting results were close would have greater importance in building a reelection strategy. We included this variable in one version of the model but found that it offers no explanatory power. Given the timing of the Chrysler reorganization—very early in Obama’s term—it seems reasonable that the administration would have chosen a “reward your supporters” strategy over a “win a swing state” strategy.

586 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

Chrysler dealership closures, however, this administration’s behavior displays politics as

usual. Our results, unlike those of the Heritage Foundation’s study, support the notion

that political considerations, not economics, influenced the decisions.

Appendix

Procedure

As noted previously, we divided the political analysis into two parts: a study at the

state level and an investigation at the individual-dealership level. First, we examined

the state level, employing an ordinary least squares (OLS) regression. The percentage

of total dealerships closed in each state served as the dependent variable in our

empirical analysis. The data appear in table 1. Independent variables include the

percentage of dealerships in each state that carry either the full line of Chrysler

products or the full line of Chrysler products plus trucks. We derived information

about the line of vehicles carried by each dealership from Chrysler’s filings with the

Bankruptcy Court.

The state unemployment rate in 2009 is included as an explanatory variable for

dealership profitability. If maximizing sales per dealership was the decision makers’

goal, then it was expected that more dealerships would be closed in states with weaker

economies. In the first OLS model, we included the total number of dealerships in the

state divided by state population in millions. In the second specification, we entered

total number of dealerships per ten thousand square miles. Both obviously measure

dealership saturation, and the expectation was that more franchise agreements would

a priori likely be voided in states with greater intrabrand competition. The court

filings provided the location of each dealership.

We included two political variables in the models. The first is the percentage of a

state’s vote cast for Obama in the presidential election. The second is the number of

electoral votes per capita in each state. OLS regression results are presented in table 2.

Next, we conducted an examination of each individual dealership. Because the

variable of interest in this case is whether a particular firm was allowed to remain open

or forced to close, a logistic model is estimated. To determine the extent to which

nonpolitical considerations influenced the closure decision, once again we included a

number of independent variables suggested by Chrysler spokespersons.

The number or “line” of Chrysler products that each dealership offered for sale

served as one such independent variable. If carrying a more extensive line of Chrysler

products was associated with remaining open, as suggested, then this variable’s

regression coefficient was expected to be negative and statistically significant. To

determine whether reducing intrabrand competition played a role in the decision to

close, we calculated the ratio of county population to the number of dealerships in the

county for each firm. A smaller ratio indicates fewer potential customers per dealer-

ship. Hence, the likelihood of closure should have increased as the ratio of potential

customers per dealership declined. Finally, we included the county unemployment

GOVERNMENT BEHIND THE WHEEL F 587

VOLUME 15, NUMBER 4, SPRING 2011

rate in the model. If economic considerations played a role in the decision to close a

dealership, this variable’s regression coefficient would be expected to be positive and

significant. High levels of unemployment should have translated into lower dealership

profitability and thus greater likelihood of termination—at least according to Chrysler

officials. Table 3 provides the results for the logistic model.

Results

The OLS models that examine the percentage of dealerships closed in each state both

explain approximately 36 percent of the variance in the dependent variable. Neither

the unemployment rate nor either of the measures of intrabrand competition has any

explanatory power. However, as the percentage of total dealerships offering a more

Table 2 OLS Regressions

Variables Model 1 Model 2

Constant 76.55 (5.56*) 81.0 (6.2*)

Obama Vote –0.298 (2.22**) –0.31 (2.34**)

Unemployment Rate –0.72 (1.24) –0.85 (1.55)

Full Line –0.39 (4.61*) –0.41 (4.41*)

Electoral Votes –1.77 (1.28) –1.48 (1.19)

Dealerships/Population 0.121 (0.57)

Dealerships/Area –0.028 (0.38)

R-squared 0.358 0.356

Note: T-statistics in parentheses.

*Statistical significance at the 1 percent level of confidence.

**Statistical significance at the 5 percent level of confidence.

Table 3 Logit Analysis

Parameter Estimate Pr > ChiSq

Intercept 3.56 0.0001*

Full Line –0.992 0.0001*

Unemployment 0.001 0.936

Population/Dealership 0.0000001 0.029**

Representative –0.148 0.127

Obama Vote –3.12 0.0001*

Note: “Representative” is equal to one if the congressperson representing the dealership’s

district is a Democrat and zero otherwise.

*Significant at the 1 percent level of confidence.

**Significant at the 5 percent level of confidence.

588 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

complete line of Chrysler products in the state increased, the percentage of dealer-

ships closed in the state decreased, and this result is statistically significant at the

1 percent level of confidence.

Electoral votes per capita has the correct sign but lacks statistical significance.

The results do indicate that a smaller percentage of dealerships was closed in states

that offered greater political support for Obama. This result is significant at the

5 percent level of confidence. Both specifications indicate that political consider-

ations influenced dealership franchise termination. In particular, the results indicate

that rather than attempting to win new voters, Obama instead used his influence

to reward his allies by protecting dealerships where his support was stronger.

The logistic regression likewise supports the service of a political motivation in

the decisions regarding which dealerships to close. The county unemployment rate

offers no explanatory power in the model. Population per dealership is significant at

the 5 percent level of confidence, but its sign is the opposite of what was expected:

Chrysler dealerships with a larger customer base were terminated, whereas firms with

a smaller base were allowed to remain open. However, the full-line variable is signif-

icantly related to closure and behaves as expected.10 Dealerships that did not offer as

many Chrysler products were more likely to lose their franchise.

Because we examine individual firms, the variable “Representative” is included.

Representative measures the party of the congressperson in the U.S. House of

Representatives in each dealership’s district. The variable is insignificant in the analy-

sis. But the independent variable that measured support for Obama in his contest

with John McCain is significant at the 1 percent level of confidence. As the percentage

of the vote for Obama increased in a state, the probability that dealerships in that state

would survive increased as well.

As noted previously, support for Obama in the presidential contest is correlated

with higher state average income. In other words, richer states offered more political

support for Obama than did poorer states. Thus, the Obama variable may be serving

as a proxy for income in our models, so that executives, in making the decision to

close dealerships, simply terminated firms in lower-income areas.

To examine this possibility, we estimated regressions that include state median

income, full line of Chrysler products, and Obama. In all the specifications, the

Obama and full-line variables behaved as before and are statistically significant. How-

ever, median income does not offer any explanatory power. In addition, we

constructed a correlation matrix with median income, Obama vote, and the percent-

age of dealerships closed. Although the Obama vote and median income are highly

correlated, median income and percentage of dealerships closed are not related.

Therefore, our evidence of political influence remains strong. The regressions and

the correlation matrix are included in table 4.

10. The full-line variable might represent higher income and more prosperous locations. The variable is not significantly correlated with income per capita or with the unemployment rate.

GOVERNMENT BEHIND THE WHEEL F 589

VOLUME 15, NUMBER 4, SPRING 2011

References

America’s Other Auto Industry. 2008. Wall Street Journal, Review and Outlook, December 1.

Anderson, Gary M., and Robert D. Tollison. 1991. Congressional Influence and Pattern of

New Deal Spending, 1933–1939. Journal of Law and Economics 34 (April): 161–75.

Becker, Bernie. 2009. GM and Chrysler Defend Dealer Closings to Senate Panel. New York

Times, June 3. Available at: http://www.nytimes.com.

Brito, Jerry, and Veronique de Rugy. 2009. Stimulus Facts. Working Paper no. 09-46.

Washington, D.C.: Mercatus Center, George Mason University, December.

Congress Plans Remedy in GM, Chrysler Dealer Dispute. 2009. Reuters, December 8.

Couch, Jim F., Keith Atkinson, Tommie Singleton, and Peter Williams. 1999. Political Influ-

ence and the Internal Revenue Service. Cato Journal 19 (Fall): 313–22.

Couch, Jim F., and William F. Shughart. 1998. The Political Economy of the New Deal. Chel-

tenham, U.K.: Edward Elgar.

Couch, Jim F., Robert J. Williams, and William H. Wells. 2008. Environmental Protection

Agency Enforcement Patterns: A Case of Political Pork Barrel? In Political Economy, Lin-

guistics, and Culture: Crossing Bridges, edited by Jürgen G. Backhaus, 233–39. European

Heritage in Economics and the Social Sciences, vol. 5. Berlin: Springer.

Faith, R. L., D. R. Leavens, and R. D. Tollison. [1982] 1995. Antitrust Pork Barrel. In The

Causes and Consequences of Antitrust, edited by William F. Shughart, 201–12. Chicago:

University of Chicago Press.

Garrett, Thomas A., and Russell S. Sobel. 2003. The Political Economy of FEMA Disaster

Payments. Economic Inquiry 41, no. 3 (July): 496–509.

Goff, Brian, and Robert Tollison. 1987. The Allocation of Death in the Vietnam War. Southern

Economic Journal 54 (October): 316–21.

Table 4 Correlations

Percentage of

Obama Vote Unemployment Rate

Percentage of

Dealerships Closed

Unemployment Rate 0.207 (0.149)

Percentage of

Dealerships Closed

–0.056 (0.701) –0.011 (0.941)

Median Income 0.481 (0.000) –0.151 (0.294) 0.065 (0.655)

Note: Pearson correlation values are in parenthesis. A number of regressions were run with

income included as an explanatory variable. Median income was never statistically significant in

any of the models, but both the full-line and Obama variables retained their statistical signifi-

cance. The results from two specifications are:

% Dealerships Closed = 54.1 - 0.283 %Full Line - 0.000181 Median Income

% Dealerships Closed = 64.4 - 0.328 %Full Line - 0.252 %Obama - 0.000072 Median Income

590 F JIM F. COUCH ET AL.

THE INDEPENDENT REVIEW

Isadore, Chris. 2008. Bush Announces Auto Rescue. Money.cnn.com, December 19.

King, Neil, and John D. Stoll. 2009. Auto Task Force Set to Back More Loans—with Strings.

Wall Street Journal, March 26.

Lord, Dennis J. 1992. Locational Dynamics of Automobile Dealerships and Explanations

for Spatial Clustering. International Review of Retail, Distribution, and Consumer

Research 2, no. 3 (July): 283–308.

Maltby, Emily. 2009. Chrysler Dealerships Fight Closings: New Law May Help Reinstate

Stores; Some Experts Call Closure Process a Mistake. Wall Street Journal, December 31.

Obama Is Interviewed by Michael Smerconish. 2009. Washington Post, August 20. Available at:

http://www.Washingtonpost.com.

Shepardson, David, and Gordon Trowbridge. 2009. Auto Task Force Taking Shape. Detroit

News.com, Washington Bureau, February 21.

Sherk, James, and James Gattuso. 2009. Closing Car Dealerships: A Matter of Economics, Not

Politics. Backgrounder no. 2296. Washington, D.C.: Heritage Foundation, July 8.

Shughart, William F., II. 2008. Public Choice. In The Concise Encyclopedia of Economics, edited

by David R. Henderson, 427–30. Indianapolis: Liberty Fund.

Tapscott, Mark. 2009. Furor Grows over Partisan Car Dealership Closings. Washington Exam-

iner, May 27.

Valdez-Dapena, Peter. 2009. Chrysler Closing 789 Dealerships. CNNMoney.com, May 15.

White House Press Office. 2009. Remarks by the President on the American Automobile

Industry. March 30. Available at: http://www.whitehouse.gov.

Wright, Gavin. 1974. The Political Economy of New Deal Spending: An Econometric Analysis.

Review of Economics and Statistics 56 (February): 30–38.

Young, Marilyn, Michael Reksulak, and William F. Shughart. 2001. The Political Economy of

the IRS. Economics and Politics 13, no. 2: 201–20.

GOVERNMENT BEHIND THE WHEEL F 591

VOLUME 15, NUMBER 4, SPRING 2011