A report released by the Indiana Chamber of Commerce last month calls for cutting Indianans’ wages and benefits as a strategy for attracting more companies to relocate to the state. Unfortunately, not only is a wage cut the last thing most Indiana families need right now, but the report’s methodology is so faulty that its numbers are highly misleading.
The particular proposal that the business lobby is promoting is a “right-to-work” law. Misleadingly named “right-to-work” (RTW) laws do not, as some unfamiliar with the term may assume, guarantee employment for those ready and willing to go to work. Rather, they make it illegal for a group of unionized workers to negotiate a contract that requires each employee who benefits from the contract terms to pay his or her share of the costs of negotiating and policing the contract. By making it harder for workers’ organizations to sustain themselves financially, RTW laws seek to undermine unions’ bargaining strength.1 When unions are weakened, wages and benefits decline for all workers—including workers who are not in a union—as competitive pressures on nonunion employers to meet union compensation standards are lessened. By cutting wages and benefits, right-to-work laws will make places like Indiana more attractive to outside investors, RTW proponents argue.
It may be ironic to hear business owners insisting that the key to economic progress is cutting other people’s wages. The bigger problem with the Chamber’s study, however, is its faulty and ideologically distorted methodology.
If a college student presented this analysis for their thesis, it would be rejected for faulty methodology. In America, anyone is free to advocate a personal ideological agenda, but both legislators and the public at large deserve to know the difference between ideological passion and scientific fact. Unfortunately, the Chamber’s study fails this most basic standard of democratic debate.
Using ‘averages’ to obscure facts
The Chamber’s argument is this: From 1977 to 2008, personal income and overall employment grew faster in the 22 so-called right-to-work states than in the 28 states that lack RTW laws. Based on these numbers, the Chamber asserts first that right-to work is the primary cause of differences in growth rates among the states; and, second, that if Indiana adopts a right-to-work law in 2011, the state will see its future income and jobs grow at a rate similar to the RTW average over the past 30 years.
At first glance, the argument, articulated in the Chamber’s study, may seem convincing. “States with RTW laws have experienced above average economic growth while states without such laws have seen below average growth,” the Chamber reports.2 Someone reading this statement might conclude that all right-to-work states enjoyed rapid economic growth, while all non-RTW states were doomed to sluggishness—that if all the states were lined up in order of growth, we would find all the RTW states up front and all others at the back of the line. In reality, nothing could be further from the truth.
Figure A shows the exact same data used by Richard Vedder, the Ohio University professor who conducted the Chamber’s study. But rather than the overall average, the growth rates for each individual state are shown, with right-to-work states shown in gray. The immediate fact that jumps out from this data is that there is no clear relationship between income growth and right-to-work laws. Indeed, the two fastest growing states were Massachusetts and Connecticut—both non-RTW states with relatively high rates of unionization. At 88.9%, Massachusetts’ growth rate was almost 50% higher than the average of all RTW states. Ten non-RTW jurisdictions (nine states plus the District of Colombia) all enjoyed income growth over this period that was greater than 17 of the 22 RTW states.
In this context, it is unsurprising that the Chamber did not publish the specific, state-by-state numbers underlying its overall average. Its report states that, on average, “growth in real per capita incomes in RTW states is substantially higher than both the national average and non-RTW states.”3 This statement is statistically true, but only in the same way that it’s true that if Bill Gates walks into a bar, everyone in the bar is suddenly, on average, a multimillionaire. The problem with averages presented in the absence of standard deviations is that they create the misleading impression that the average is more or less representative of everyone in the group.
In fact, by reporting only the average growth rate for right-to-work states, the report obscures huge disparities among RTW states. For instance, in 1977-2008, per capita income grew by 82% in North Dakota but by less than half as much (32.5%) in Nevada. Indeed, Nevadans’ income grew more slowly than either the national average or the average of non-RTW states. The average of 22 very diverse states reveals nothing about the economic fortunes of any particular state. The residents of Nevada are no more helped by being associated with the right-to-work average than my personal wealth would be if Bill Gates walked into my favorite watering hole.
But this fact points to a glaring failure in the calculations of Vedder and the Chamber of Commerce. Vedder contends that a so-called right-to-work law—in and of itself, independent of any other state policy or economic factor—would increase any state’s economic growth by more than 20%, and would have increased income by nearly $12,000 per year for a family of four in Indiana. This is a far greater impact than policy analysts claim for any other economic development strategy. Indeed, the Chamber claims that a RTW law is “the single most important step Indiana lawmakers could take in putting more Hoosiers back to work.”4 But if right-to-work is the key to economic growth, how do Vedder and the Chamber explain the fact that four of the top five states are non-RTW? If labor laws determine income growth, shouldn’t Nevada and Mississippi and Iowa model themselves on Massachusetts? For that matter, if right-to-work laws are the primary factor determining income, how do Vedder and the Chamber explain the gap between high-growth and low-growth RTW states?
Confronted with these contradictions, Vedder backtracks. “Obviously, the model was far from perfect,” he concedes. “No doubt there are some admitted error and bias problems.”5 But such concessions are nowhere in the published report, nor in the Chamber’s communication with legislators.
The inability of right-to-work laws to explain economic facts is reinforced by the data on job growth. As with personal income, the Chamber trumpets the fact that, over the past 30 years, job growth has been faster, on average, in states with RTW laws. Once again, the Chamber did not publish the actual state-by-state numbers; and these data reveal the same contradictions as with income growth. For example, employment in 1977–2008 grew by 105% in non-RTW New Mexico, more than 70% faster than its RTW neighbor, Oklahoma. Indeed, the non-RTW states of Colorado, Washington, New Mexico, and New Hampshire each saw jobs grow faster over this period than 15 of the 22 RTW states.6 And by the end of 2010, both the highest and the lowest state unemployment rates were found in right-to-work jurisdictions.7
Finally, while the Chamber report claims that the presence or absence of state right-to-work laws explains both job growth and income growth, these two trends are not in synch. For instance, while Nevada had the single highest job growth of any state from 1977 to 2008 (322%), it ranked 48th in income growth (32.5%).
Thus, for the Chamber’s theory to be correct—i.e., that right-to-work is the primary determinant of job and income growth—it would have to simultaneously explain why one RTW state grows six times faster than another; why four of the five fastest-growing states are non-RTW; and why right-to-work Nevada had both the highest job growth and nearly the lowest income growth in the same period. The Chamber makes no attempt to explain these contradictions.
The simple truth is that if states with so-called right-to-work laws can experience either dramatic growth or steep decline, and if both RTW and non-RTW states can foster booming job markets, something in these states’ economies, demographics, or policies other than RTW clearly are driving their growth.
Disguising correlation as causality
Beyond the misleading nature of its averages, Vedder’s report suffers from another fundamental failure: It asserts that because one phenomenon is statistically associated with another, the first must be the cause of the second. In the Vedder report and elsewhere, the Chamber frequently argues that because average job growth has been faster in the 22 right-to-work states than in the rest of the country, right-to-work must be the cause of this success. This suggestion ignores a fundamental tenet of statistics: the difference between correlation and causality.
It’s a big world and if you look hard enough, it’s easy to find inauthentic correlations.8 From 2000 to 2009, for example, average job growth was nine times higher in states with names that start with the letters N through Z than in states with names that begin in the first half of the alphabet.9 Yet no one would suggest that Indiana could improve job growth by changing its name to “Tindiana.”
More to the point, if simple correlation is sufficient grounds to argue for policy changes, then there is plentiful data pointing in the opposite direction from the Chamber’s conclusions. It does not necessarily make sense, for instance, to compare Indiana with the average of all right-to-work states—a group dominated by sunny locales with cheaper real estate, and dominated above all by Texas with its large oil and gas industry, which Indiana cannot hope to replicate, no matter what laws are enacted. Rather than comparing Indiana with the RTW average, it might be more logical to compare it with the RTW state that is closest to Indiana’s economic region—Iowa. Looking at job growth, we see that Iowa’s employment grew by 34.4% from 1977 to 2008, while Indiana’s grew by 42.8%. If we assume that correlation is causality, we would conclude that Iowa can increase its job growth by repealing right-to-work and making itself more like Indiana.
Similarly, Vedder forcefully argues that higher average income growth in the right-to-work states means that other states should adopt similar statutes in order to spur economic growth. But this logic can also lead in the opposite direction. Thus, the Bureau of Labor Statistics reports that per capita income in the United States increased by 77% from 1977 to 2008, but per capita income in Scandinavia grew by 92.4% over the same period (see Figure B).10 Not only do none of the Scandinavian countries have right-to-work laws, but the share of their employees represented by unions is much higher than anyplace in the United States, ranging from a low of 72% in Norway to a high of 92% in Sweden. Would the Chamber argue that Scandinavians’ superior growth indicates that we should abolish right-to-work and encourage dramatically increased unionization as a path to economic revival?
The correct way to measure economic impacts: statistical controls
Clearly, what both sides of this debate must aim to discover is how right-to-work laws impact a state’s job growth, all other things being equal. The methodological difficulty lies in defining what “all other things” means. Noticing, for example, that one state has experienced faster growth than another, one might wonder if the difference is due not to labor laws, but to the difference in the educational level of the workforce, the degree of urbanization, the industries and occupations that dominate the economy, the cost of real estate, the presence of natural resources, personal and corporate tax rates—the list lengthens very quickly.
It is notoriously difficult to isolate the economic impact of a single state policy. But it is incumbent on economists to do everything possible toward this goal before reporting “results” to public officials. Unfortunately, the Chamber’s study fails this standard as well. Table 1 below compares the range of factors controlled for in a recent analysis by the Economic Policy Institute of the impact of right-to-work laws on wages with those factors considered in the Chamber’s Indiana study. The Chamber study failed to account for such basic factors as the age, education, gender, ethnicity of the workforce, and the concentration of jobs in various industries and occupations. As a result, the chamber’s numbers, while dramatic, are wholly unreliable.
What does the research on right-to-work actually show?
The impact of right-to-work laws has been debated for decades, and scholars have produced a number of scientifically rigorous studies on the topic. A recent review of this scholarship finds that, as statistical methods have gotten more sophisticated, and analysts have come closer to holding “all other things equal,” right-to-work laws have been shown to have no impact whatsoever in boosting a state’s job growth.11
The failure of right-to-work laws to increase job growth is particularly evident in the case of Oklahoma, the only state to have newly adopted such laws in the past 25 years. Because Oklahoma is the sole state to have enacted a right-to-work law in the post-NAFTA period—when manufacturers seeking low-wage laborers are more likely to look to Mexico or China than the Southern United States—its experience is a particularly useful guide to those states now debating such policies.
Vedder himself has elsewhere agreed that the reality of globalization makes right-to-work laws a less meaningful attraction for employers seeking low-wage labor. “I would be cautious in the future of assuming it would have the same weight” as in the past, he notes.12 Yet this qualification is not included in the Chamber’s report. On the contrary, the Chamber explicitly argues that if Indiana adopts right-to-work, its income over the next 10 years will grow at the same rate that RTW states averaged over the past 30 years.13
Unfortunately, Oklahoma experienced no improvement in its unemployment rate after passing right-to-work; its manufacturing sector shrank dramatically, and the number of new companies coming into the state fell by one-third in the decade following adoption of the labor statute. Thus, recent evidence points strongly to the conclusion that misleadingly named right-to-work laws are impotent as a strategy for increasing employment.
At the same time, careful statistical analysis of the impact of right-to-work laws on wages suggests that such laws are indeed successful in driving down wage and benefit levels (by approximately 3%)—for union and nonunion employees alike.14
Politics and policy
If right-to-work policies have been shown to lower wages while failing to increase employment, why would anyone promote them?
Right-to-work proponents sometimes claim they are motivated by a principled commitment to personal liberty. Vedder, for instance, suggests that the goal of right-to-work is “greater personal liberty with respect to employment.”15
But neither Vedder nor the Chamber advocates a broader right of persons to be free from the requirement to pay dues to organizations not of their choosing. All attorneys who appear in court in the United States, for instance, must be dues-paying members of the Bar Association. And many residents must pay dues to the homeowners’ association for the residential community in which they live. The Chamber has never advocated outlawing such practices.
Even within the world of employment, the so-called “right-to-work” that Vedder and the Chamber promote is exceedingly narrow. For instance, it does not include the right of those who report fraudulent workplace activity to work free from fear of retaliation—indeed, the Chamber has opposed increased protections for whistleblowers. It does not even include a right for employees to remain on the job rather than spending time against their will attending sessions advertising their employer’s views on religion, politics, or union issues; the Chamber went to court to block a law that would have provided this protection.16 If the only right of concern to right-to-work proponents is the right to undermine workers’ collective strength by making unions financially insecure, it’s hard to conclude that this passion stems from some broader commitment to employee rights.
Clearly, some employers seek to limit their employees’ ability to bargain for wages and benefits out of corporate financial interests. Likewise, “right-to-work” is obviously championed, in part, by those who are ideologically opposed to unions.
Prof. Vedder certainly fits this latter category. Vedder is not only notoriously anti-union, he insists that inequality shouldn’t be so troubling because “comparing the rich and the poor, it is worth noting that the rich work a lot more.”17 He suggests that the Great Depression was caused by workers’ wages being too high,18 and has argued against both the minimum wage and the eight-hour day, proclaiming that “the eroding of employment liberty” began with “laws restricting hours and setting minimum wages.”19 The Chamber opposed the minimum wage and the 8-hour day, too, just as it opposed the federal Occupational Safety and Health Act, the Family and Medical Leave Act, and the more recent Lilly Ledbetter law to prevent sex discrimination in the workplace.
Both Vedder and the Chamber are entitled to their opinions. But the people of Indiana, and especially the legislators charged with charting the state’s economic future, are entitled to know the difference between passionate ideology and scientific fact. By ignoring this distinction, the Chamber has done a disservice to lawmakers and to the public.
—Gordon Lafer is an Associate Professor at the Labor Education and Research Center at the University of Oregon.
His work concentrates on strategic planning, strategic research, and labor and employment policy issues.
- As the Chamber report notes, “the goal of labor unions [is]to increase wages and benefits for their members,” and the goal of right-to-work is to cut wages in the hopes of attracting outside investors. Richard Vedder, et al., Right-to-Work and Indiana’s Economic Future, Indiana Chamber of Commerce Foundation, January 2011, pp. 6-7.
- Richard Vedder, et al., Right-to-Work and Indiana’s Economic Future, Indiana Chamber of Commerce Foundation, January 2011, p. 12.
- Vedder, et al., p. 11.
- Mike Blakley, 2011 Chair of the Indiana Chamber of Commerce, quoted in Chamber press release issued with publication of the Vedder report.
- Prof. Richard Vedder, conversation with the author, Feb. 14, 2011.
- U.S. Bureau of Economic Analysis, Total Employment by State, 1977-2008. source:www.bea.gov/regional/spi
- In November 2010, the highest state unemployment rate was Nevada’s, at 14.3%, while the lowest was North Dakota’s, at 3.8%. Reported in Oklahoma Economic Indicators, December 2010, Oklahoma Employment Security Commission, http://osec.ok.gov, accessed Jan. 21, 2011.
- Indeed, Purdue University in Indianapolis has even hosted a “spurious correlations contest” that discovered, among other things, that ministers’ salaries are statistically correlated with the price of vodka. http://www.morris.umn.edu/~sungurea/introstat/public/instruction/causation.html
- U.S. Bureau of Labor Statistics, States and selected areas: Employment status of the civilian noninstitutional population, 1976 to 2009 annual averages.
- U.S. Bureau of Labor Statistics, Division of International Labor Comparisons, October 21, 2010.
- See Gordon Lafer, Does “Right-to-Work” Create Jobs? Answers From Oklahoma, Economic Policy Institute, Feb. 28, 2011.
- Conversation with the author, Feb. 14, 2011.
- Indiana Chamber of Commerce, “New Right-to-Work Study, Statewide poll Shows Policy Benefits,” January 31, 2011. Accessed Feb. 14, 2011 at www.indianachamber.com/index.php/chamber-releases-right-to-work-study-statewide-poll.
- Elise Gould and Heidi Shierholz, The compensation penalty of “right-to-work” laws, Economic Policy Institute, Feb. 17, 2011.
- Richard Vedder, “Right-to-Work Laws: Liberty, Prosperity, and Quality of Life,” Cato Journal, vol. 30, no. 1 (Winter 2010).
- Associated Oregon Industries and Chamber of Commerce of the United States of America, Complaint for Injuctive and Declaratory Relief, U.S. District Court, District of Oregon, 28 USC Sec 2201-2202; FRCP 65, 2009.
- Richard Vedder, “Economic Growth, Economic Justice, and Public Policy,” testimony to Joint Economic Committee, Jan. 31, 2007. Accessed Jan. 31, 2011 at http://www.aei.org/print?pub=speech&pubId=25553&authors=%3Ca%20href=scholar/113%3ERichard%20Vedder%3C/a%3E.
- “We estimate that in 1930 wages were 8% higher than they would have been normally given the economic conditions of the time. This devastated profits and corporate balance sheets…” Address to Council on Foreign Relations.
- Richard Vedder, “Right-to-Work Laws: Liberty, Prosperity, and Quality of Life,” Cato Journal, vol. 30, no. 1 (Winter 2010).