Will Labor Fare Better Under State Labor Relations Law?

Will Labor Fare Better Under State Labor Relations Law?
Richard B. Freeman
LERA Meetings
January 2006
Acknowledgements: I benefitted from discussions with Joel Rogers and on joint work that will appear as R.
Freeman and J. Rogers, “The Promise of Progressive Federalism,”in Making the Politics of Poverty and Inequality,
ed. Jacob Hacker, Suzanne Mettler, and Joseph Soss (New York, NY: Russell Sage Foundation, forthcoming)
This paper gives an affirmative answer to the title question. I argue that labor would fare better if the US
reduced federal preemption of private sector labor relations law and devolved the legal regulation and enforcement
of the right of association and collective bargaining to the states.
The claim that labor would do better under state law than federal law runs counter to the standard story in
US labor history and the traditional views of pro-labor scholars. US history warns against trusting states to protect
labor rights. The southern states made slavery legal until the Civil War and used state laws to suppress blacks for
over a century after the 14th Amendment. Given the chance to enact legislation harmful to unions by the Taft-Hartley
Act, 21 states decided in favor of right to work legislation that forbids firms and unions from including union
security clauses in their contracts. The archetypical villain in union organizing is a racist Southern sheriff who
abuses his legal powers to put down the Norma Raes of the world. By contrast, labor histories often credit the
National Labor Relations Act for the growth of unionism in the US in the late 1930s through the 1950s. 1 Congress’s
enactment of Title IV of the Civil Rights Act of 1964 dealt a major blow to discrimination against blacks,
particularly in southern states that would never have passed anti-discriminatory legislation on their own. From the
1960s through 2000, Congress enacted a host of other labor bills -- protecting different groups against
discrimination, regulating occupational health and safety, insuring employment retirement plans, etc, so that as of
2005, the Department of Labor administered and enforced more than 180 federal laws covering 125 million
In the area of labor relations law, on which this paper focuses, courts have held that the federal legislation
preempts state activity. States are free to legislate wages above the federal minimum and enact and enforce
employment law, which covers individuals rather than groups, but are precluded from taking independent action on
private sector labor relations issues. Much of the impetus for federal preemption came from liberal Supreme Court
justices, influenced by pro-labor legal scholars such as Archibald Cox, who argued for federal preemption although
there is no explicit Congressional statement in this regard.3 Indeed, since the Supreme Court decision in Garmon
(1959) courts have held that the NLRB has exclusive jurisdiction over conduct regulated by the LRMA. More
recent legal decisions have extended the preemption doctrine to efforts by states to limit the anti-union activity of
Even after passage of the Wagner Act, however, many workers unionized without using the Act. Freeman (1998).
Michael H. Gottesman, “Rethinking Labor Law Preemption” Yale Journal on Regulation 7 (Summer 1990): 355410
their contractors. At the same time, the Supreme Court has interpreted state sovereign immunity to bar suits for
damages by state employees for violations of federal employment rights.4
In recent years, however, labor has fared poorly under national regulations. From the 1960s through the
2000s, unions gained miniscule numbers of members through the NLRA election process. The Pension Benefit and
Guaranty Corporation which insures private defined benefit pensions has run large deficits as the private system
lurches from crisis to crisis. Relative to average hourly earnings, the federal minimum wage has fallen to its lowest
level in decades. It is barely half the minimum wages in the UK and in Ireland, despite the US having 40 percent
higher GDP per capita than those countries.
My case that labor would do better under state law begins with the standard federalism arguments for
devolving authority: that it allows experimentation in policies and creates space for divergent policies to match
citizen preferences. I show that state law has produced wider cross-state variation in the unionism of public sector
workers than federal law has produced in unionism of private sector workers. Associated with this is a higher level
of unionisation in the public sector than in the private sector. I use evidence on the variation in state policies toward
labor broadly to suggest that if states had the right to enact or enforce labor relations, some states would enact
policies more favorable to collective bargaining than the Congress, while others would enact less favorable policies.
Because private sector collective bargaining and unionisation is exceptionally low in the states likely to introduce
legislation unfavorable to collective bargaining, such legislation cannot have much effect on union density or
coverage. By contrast, a more friendly legal environment in the states that look favorably on collective bargaining
has the potential for giving workers in those states the representation and participation at the workplace that they say
they want. Finally, I argue that state legal enactment offer the best opportunity for the country to experiment with
new forms of worker organizations, including variants of the company sponsored workplace organizations that the
LMRA outlaws, to open the market for representation and participation to meet the desire of workers who want and
cannot get unions under natiional law and of workers some nonunion organization short of unions to represent them
to employers.5
Landau, Brent W. “State Employees and Soveriegn Immunity: Alternatives and Strategies for Enforcing Federal
Employment Laws” Harvard Journal on Legislation, vol 39, pp 169-212
Every survey shows that millions of nonunion workers want unions to represent them but are unable to gain that
representation. The 1994-95 Workplace Representation and Participation Survey and recent Hart Research
Associates surveys show that even larger numbers of non-union workers want employee organizations that fall short
If my analysis is correct, the union movement would do better to spend its resources to press the Congress
and courts to reduce or eliminate federal preemption of private sector labor law and to improve labor regulation and
enforcement at the state and local level than to push for “labor law reform” at the national level.
1. The Logic for Devolving Labor Relations Law to states
“ It is one of the happy incidents of the federal system that a single courageous state may, if its citizens
choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the
country”. Justice Brandeis6
The classic argument for federalism has three parts.
The first, as given by Brandeis above, is that federalism allows for experimentation so that the country can
better assess the impact of different policies and then choose the one that works best in accomplishing the public
purpose. States are well-suited to serve as laboratories of experiment because they are numerous, have boundaries
that cover populations with similar attributes and views, and have the large scale necessary for assessing major
policy initiatives. As a practical matter, policy changes are easier to achieve in a state than across the whole
country, and reduce the risks inherent in any new initiative turning sour from the country to the state. The natural
experiment where state A chooses policy I and neighboring state B chooses policy II can provide information on the
aggregate effect of the policies which elude smaller random assignment studies.
The second part of the argument for federalism is that states will imitate the most successful policy
initiatives, so that good practices spread rapidly. The competitive market moves toward the right output in part by
the exit and entry of consumers or firms. If, by chance, a given firm produces a product that consumers prefer to
others, consumers will switch to it, and other firms imitate the lucky producer or go out of business. This is natural
selection at work in the economy. Since states that choose the wrong policies do not go out of business, the market
for policies operates somewhat differently. Citizens have the choice of moving to a jurisdiction with more desireable
policies, or of taking political action to induce their jurisdiction to select their desired policies.
Whether these forces operate effectively in any reasonable time period is unclear. The long association of
states rights with slavery and disenfranchisement of blacks has made many Americans uneasy about trusting state
of a collective bargaining union. See Freeman and Rogers, What Workers Want, Cornell, 2 nd edition, forthcoming,
2006, chapter 1).
Brandeis, dissenting, New State Ice Co. v. Liebmann, 285 U.S. 262, at 311 (1932). Greve critiques the case to
which Brandeis dissented www.aei.org/publications/pubID.12743/pub_detail.asp.
power in the political area. Conservatives often express unease about state fiscal responsibility – will states will run
up huge deficits that require a national bailout or possibly default? (Ferejon and Weingast) Liberals worry that the
potential mobility of capital will lead states into a race to the bottom, as each state reduces its protections for labor to
attract investments. After all, don’t states compete for particular investments through tax holidays and subsidies that
benefit the investor more than citizens? (Donahue)
But the politics could go the other way. Citizens in a state with worse labor relations policies might see
how well people are doing in a state with better labor regulations, and move to the more desireable state or campaign
for their state to copy the preferable legislation. Federalism in welfare services and medical services for the poor
has not led to any race to the bottom among the states. Because states deliver services to citizens subject to hard
budget constraints, governors and legislators tend to be less ideological and divisive than representatives and
senators in Washington. State governments choose policies that make pragmatic sense even when those policies run
against the prevailing ideology in their party. In 2004, the Republican legislature in Nebraska, for instance, chose a
defined benefit state pension system over a defined contribution system because of evidence that this would save
taxpayers money, while the Bush Administration was pushing for private accounts to replace Social Security.
Historically, many of the basic social insurance, worker rights, public health, administrative and government
reforms adopted by the New Deal were first anticipated and developed in states, such as Wisconsin. These policies
spread to other states, and then became national law.
The third part of the argument for federalism is that in a large diverse country, federalism allows each area
to find the policies that fit it best. Mississippi gets the policies that it wants. Minnesota gets the policies it wants.
California does whatever California wants. New York is New York. And so on. Forcing every state to follow the
same rules and regulations necessarily reduces well being. In standard economic models, moreover, high
geographic mobility increases the gains from policy diversity, since people with preferences that differ from those in
their locale can move to a different locale. This is the Tiebout solution to the optimal determination of state and
local expenditures. Citizens opposed to the labor regulations in their state vote with their feet by moving to states
whose policies satisfy their preferences. Geographic mobility is the key market force, comparable to consumer
choice in a competitive market.
The conceptual case for federalism is a strong one, which has attracted increasing academic support in the
past decade.7 Still, there are potential benefits to national legislation – incorporation of externalities across state
borders, maintenance of a single market, economies of scale in setting policy -- that argue in the other direction. In
addition, federal law can spread best practice policies to states that might lack the political will or funds to protect or
help their own working citizens. Given that there are plausible arguments on both sides of the debate, which may
weigh more heavily in some situations than in others, the question of whether labor will do better under state law
cannot be answered in the abstract. It requires evidence on how states have actually regulated labor relations when
they have authority over labor laws or policies; and of the impact of their actions on the ability of workers to gain
the participation and representation at their workplace that they want.
2. Variation in State Labor Relations Law
In one respect, the US labor relations system is well-suited for examining the effect of state as opposed to
federal regulation of labor relations. This is because states have authority over labor relations for state and
municipal workers. Federal legislation sets and enforces the rules for private sector workers but state legislation sets
and enforces the relations laws for public sector workers. This means that every state has private sector workers
covered by federal law and public sector workers covered by state law, possibly in the same occupations or
activities. Since state law can vary greatly, this provides a “natural experiment” in which we can contrast outcomes
for the treatment group, public sector employees covered by that state’s public sector laws, against the control group
of a private sector employees in that state covered by the ubiquitous federal law.
To see how states regulate the state and local public sector workers over whom they have jurisdiction, I
have examined the Kim Rueben update of the NBER Valletta-Freeman state public sector labor law data set
(http://www.nber.org/publaw/). This data set categorizes public sector labor laws from 1955 to 1985 for five groups
of workers (state employees, local police, local firefighters, local teachers, other local employees) in tems of their
legal treatment of collective bargaining and the right to strike. Rueben has updated the data set through 1996.
Valletta and Freeman found huge variation in public sector labor laws among states and smaller variation among
groups of workers within states. They showed that the legal environment changed greatly in the 1960s and 1970s
See the Boston Federal Reserve New England Economic Review, June/July 1998
when many states shifted from meet and confer laws to laws that either explicitly or implicitly required government
employers to bargain with unions.
Table 1 summarizes the legal status of public sector collective bargaining in the 50 states in 1996, the last
year of the Rueben update.8 The table places the laws governing state/ occupation groups into 3 broad categories:
having a favorable legal environment for public sector collective bargaining – where laws explicitly require or
imply that public sector employers have a duty to bargain with unions; an intermediate environment where the law
requires public sector employers to meet and confer with unions but not to bargain with them; and an unfavorable
legal environment, where the law either explicitly outlaws bargaining or contains no provision for bargaining.
According to the table, sixty-four percent of the state occupation groups have laws favorable to collective
bargaining. The bulk of these observations are found in the twenty-seven pro collective bargaining states listed in
the table. Fourteen percent of the observations occur in state occupation groups covered by meet and confer laws,
largely in the five states listed in the table. Finally, 22% of the observations have laws that are unfavorable to
collective bargaining. These are concentrated in 17 states, many in the South. Since 1996 the legal changes
governing state employees have been modest, though there are notable developments in particular states. 9
The states whose legal stance is unfavorable to public sector collective bargaining overlap to a considerable
degree with the states that have outlawed union security clauses in private sector collective bargaining. Twenty one
states have enacted such right to work (RTW) laws. 10 Fifteen states with RTW have public sector labor regulations
in the “least favorable to collective bargaining” category. Two RTW states have regulations that fit into the
intermediate “meet and confer” grouping. Four have regulations that are favorable to collective bargaining. 11 In
percentage terms, 71% of the right to work states are in the least favorable category for public sector collective
For the most recent data on state regulation of teachers see Krueger, Carl, State Collective Bargaining Policies for
Teachers http://www.ecs.org/clearinghouse/37/48/3748.pdf
Alabama, Arizona, Arkansas, Florida, Georgia, Idaho, Iowa, Kansas, Louisiana, Mississippi, Nevada, Nebraska,
North Carolina, North Dakato, Oklahoma, South Carolina South Dakota, Tennessee, Texas, Utah, Virginia
Here is the cross tabulation of RTW laws by the legal environment for public sector bargaining:
Public Sector Legal Environment toward collective bargaining
with the definition of public sector legal environment as in table 1.
bargaining. On the basis of their actions, I would expect that if the states that have legislation unfavorable to public
sector collective bargaining and RTW laws were to regulate private sector labor relations, they would create an
unfavorable legal environment for unions and collective bargaining in the private sector as they have done in the
public sector. On the other side of the divide, 72% of non RTW states are in the most favorable category for public
sector collective bargaining, which suggests that they might institute more favorable regimes for collective
bargaining than the national government has done.
How have the different public sector labor regimes affected the prevalence of collective bargaining in a
given state?
To see the extent to which differences in public sector labor are related to public sector unionization and
collective bargaining, I have linked the public sector labor law measures to estimates of the percentage of public
sector workers covered by collective bargaining from the CPS, as given by Unionstats.com
(http://www.unionstats.com). Panel A of Table 2 compares public sector bargaining coverage and private sector
bargaining coverage among states with different public sector bargaining laws. The figures in the line public sector,
mean coverage, show that states with favorable laws to collective bargaining have more than twice the coverage as
states with unfavorable laws, while those with intermediate laws have coverage rates just modestly higher than those
with unfavorable laws. The figures in the line private sector, mean coverage, show much smaller differences in
density for the private sector. The figures in the line within-state differences give the difference between the public
sector density and private sector density by state. States with favorable public sector collective bargaining laws
have statistically significant higher differences between public and private density.
The regressions in panel B of table 2 exploit the within state variation in unionization and laws to examine
whether state laws more favorable to collective bargaining are associated with greater collective bargaining
coverage in 2004, conditional on private sector unionism in the state and the presence of a Right toWork Law, using
a linear and a log-linear regression model. Line 1 gives the univariate regression of public sector coverage on
dummy variables for whether the state had public sector laws that were either favorable or unfavorable to collective
bargaining; the deleted group are the states with intermediate laws. It replicates the mean differences in panel A in
showing the huge impact of favorable laws on collective bargaining coverage in the public sector. Line 2 adds the
level of private sector bargaining and whether or not the state has a right to work law as additional independent
variables. If public sector bargaining was due to the state’s general attitude toward unionism, as reflected in the
private sector unionization rate and decision to enact an RTW law, the public sector favorable law variable would
lose its significance. It does not. The level of private sector coverage has a sizeable significant effect on the percent
covered, while the presence of a right to work law has a moderate negative effect.
Lines 3 and 4 transform the coverage variables into ln form in order to get an estimate of elasticity of coverage to
having favorable or unfavorable labor laws. The coefficient on the favorable law variable is the 0.50 in line 4. This
suggests that moving from an intermediate legal status to a favorable status for collective bargaining raises coverage
by 0.50 ln points or about 65% (= exp 0.50).
To be sure, this cross section relation does not establish causality. If Missisisippi enacted a public sector
labor law comparable to Wisconsin’s, its public sector might not have the same level of collective bargaining as
Wisconsin. But the cross section creates a presumption of some effect. And studies of public sector unionization
before and after changes in state labor laws in the 1970s-1980s suggest that the laws had an independent effect on
unionization, or at least on the timing of union growth in the public sector (see Farber, 1988, Saltzman,
1988,Valletta and Freeman, 1988). Farber 2005 uses the within-state variation in laws by type of worker and finds
that union density is signicantly higher where unions are allowed to negotiate union security provisions (e.g., agency
shop) and where employers have a legal duty
to bargain with labor unions.
Finally, under the standard argument for federalism, the variation in state regulations of public sector
bargaining ought to produce greater dispersion in the percentage of state and local employees covered by collective
bargaining than the same federal law would produce in the percentage of private sector employees covered by
collective bargaining. In 2004 the standard deviation of public sector union coverage rate across states was 16.5
whereas the standard deviation of private sector union coverage was 3.8. But since the public sector rate of
unionization was much higher, its standard deviation was almost certainly higher as well. A tougher test of the
claim that variation in state regulations produces the greater variation in coverage in the public sector than in the
private sector is to compare the standard deviation of each sector with the standard deviation from a binomial
distribution, with the mean rate of coverage across states taken as the probability of coverage. For the public sector
the actual standard deviation is 2.44 times the expected standard deviation. while for the private sector, the actual
standard deviation is 1.00 times the expected standard deviation. Thus, the public sector shows statistically
significantly more variation than one would expect from the binomial model while the private sector shows the same
variation one would expect from the binomial model. The only plausible explanation is the difference in legal labor
law regimes.
To be sure, there are reasons for public and private sector union density to diverge even under the same
legal regime. The profit motive ought to induce greater opposition to unions among private sector employers than
among public sector employers, for whom union members and their allies can be an important part of the electorate.
Workers may have different desires for unions in the two sectors, depending on lengths of employment, civil service
regulations, and market or political pressures. Still, the difference in the legal regimes has been associated with
trends in collective bargaining coverage that have produced huge differences in the level of bargaining between
workers covered by state regulations and those covered by federal regulations. As of 2005, private sector union
density has fallen to 7-8% of the work force, while public sector density has stabilized at about 35% of the public
sector work force (see Farber, 2005 for a detailed analysis of the changing fortunes of unions in the two sectors). As
table 2 shows, the gap between the public sector and the public sector is greatest in states with favorable public
sector labor laws. Since private sector unionism declined in virtually all states at roughly similar rates in the 1980s
and 1990s, while public sector density has stabilized, the gap between the public/private sector increased most in the
states with favorable public sector labor laws.
Finally, the experience of Canada with provincial determination of much labor legislation is also consistent
with the view that organized labor does better when labor laws are set at lower levels. Union density is falling in
Canada, but much more gradually than in the US. At various times provinces have enacted card check legislation
that US unions would regard as dream “pie in the sky” to get from the US Congress. Others have enacted less
favorable legislation. There are sufficient changes in legislation in the same province over time to show that the
provincial legislation affects unionisation rates.
3. Variation in Other Labor Regulations
That organized labor does better in the public sector with state laws governing state and local workers than
it does in the private sector with national laws governing private sector workers is consistent with the claim that
labor would do better under state regulations. But it does not establish the claim since, as noted, there are other
factors that differentiate public and private sector employer and employee behavior. To assess the plausibility of the
claim further, I examine next state regulations of private sector labor workers in areas where states have authority.
If there is considerable variation in state laws governing private sector labor and if those laws are more favorable to
labor in states that have public sector regulations favorable to collective bargaining, it would seem reasonable to
expect that those states would create a legal environment favorable to private sector unionism if they had authority
over the private sector. By the same logic, the opposite would be true for states that had unfavorable public sector
regulations and enacted laws less favorable to private sector labor in areas of state jurisdiction. Devolving private
sector labor relations law to the states would thus produce legal environments that would make private sector
collective bargaining easier in some states and more difficult in others, so that the key question becomes what
happens on average among the states.
Table 3 gives the evidence that there is considerable variation in state laws governing private sector labor
and that the laws are more favorable to labor in the states that have also enacted favorable public sector collective
bargaining laws. The table focuses on eight laws covering different aspects of labor regulation in the private sector,
ranging from minimum wages to transitional food stamps for families leaving welfare. 12 In each case, having the
regulation means that the state is more favorable to workers. To keep the statistics consistent, I coded the two
measures that are naturally numeric – the criterion for medicaid eligibility, and support for need based aid for
postsecondary students – as dichotomous variables around the cut point specified . There are other measures of state
legal and administrative activity that impact workers, such as the existence and operation of state equal opportunity
offices, the activities of state departments of labor, workmen’s compensation, and so on, that might be added to the
Column 1 of the table records the number of states which have adopted each of the particular policies,
while columns 2-3 show the number of states with the policy in states that have favorable, and
intermediate/unfavorable public sector bargaining laws. For ease of presentation I have collapsed the intermediate
and unfavorable states together. The percentages in each box measure the percentage of states in that category that
had the particular law in the row. For instance, the line minimum wages above federal level shows that 16 states had
above federal minimum wages in 2005 and that all 16, or 100%, were in states with favorable public sector labor
laws. Since there were 27 states with favorable public sector labor laws, 59% of them had the law whereas 0% of
The eight laws exceed the number given by PERI in their analysis of how states treat labor or the Fraser Institute
indicator of labor market freedom,. Both of these indicators mix laws with outcomes from other policies.
the states in the intermediate/unfavorable group had minimum wages above the federal level. The next line 2 shows
that 15 states offered medicaid to workers with earnings over 100% of the poverty line and that 13 were in the states
with favorable public sector labor laws while 2 were in intermediate/unfavorable group of states. This gives
probabilities of having that law of 48% and 9%
By way of summary, at the bottom of the table I give the total number of the laws in each group of states
and the average number of laws adopted in each group. The states with favorable public sector collective bargaining
laws averaged 4.3 laws compared to 2.2 laws for the other states. The differences between the favorable and
intermediate and unfavorable states in private sector labor regulastions is statistically significant.
Finally, looking at the entire distribution of pro labor laws/policies by state, I counted the number of prolabor laws/policies that each state had and compared that count to the distribution that would have resulted if the
states had drawn their policies/laws from an urn in which the probability of a pro-labor policy was the probability of
having a law, 45% (= 179/400, the total number of possible adoptions of a labor-favorable law). The actual
distribution differed greatly from what one would expect if states randomly, with much greater numbers of
observations in the tails. For instance, Wisconsin and California had most of the laws/ policies, while several had
one or none of the policies.
In sum, states policy stances toward private sector labor vary considerably in ways associated with the way
they regulate public sector collective bargaining. Some states are favorable to labor interests and likely to enact or
enforce laws favorable to labor in the private sector as they do in the private sector labor domains over which they
have authority. On the basis of their legal stance in the public sector and pro labor legislationi in the private sector, I
would expect them to create favorable legal environments for collective bargaining in the private sector. Other
states are less favorable to labor interests in public sector collective barganinng law and in laws or policies for labor
in the private sector. They are likely to create a more hostile legal environment for collective bargaining than under
federal legislation. How these two forces would work out for the country as a whole depends on the nature of the
shift from federal to state law and the numbers of workers in the groups of states.
4. The Net Outcome
There is one way in which devolving labor law from the federal government to to state government would
guarantee that labor gains. This would be if the devolution follow edthe Taft-Hartley model of allowing states to go
beyond federal rules in one direction only. Just as the Taft-Harley law allows states to outlaw union security
provisions but does not give them any way to enact laws to help protect workers rights of association or to
strengthen collective bargaining, Congress could give states the right to pass laws or enforce existing law onlyin
ways that enhanced worker rightscollective bargaining.
Some states would surely act on this. In the 1980s, Wisconsin directed its Department of Industry, Labor
and Human Relations to list firms who violated the LMRA over a five year period and to bar those firms from state
procurement. In 1986 the Supreme Court unanimously ruled that by prohibiting state purchases from repeat labor
law violators, the state was acting as a regulator rather than as a purchaser and held that the LMRA preempted the
Wisconsin statute.13 In 2000 California tried to get around this rule by enacting a law that prohibited employers
from using funds received by the state to oppose union organizing. 14 The emphasis was on the state’s rights as a
purchaser rather than as an entity seeking to enforce the LMRA . In 2002 New York enacted similar legislation, and
in 2003 the Hawaii AFL-CIO Federation proposed a similar statute for Hawaii. Employers challenged the California
law on the notion that it limited the legal rights of employers under the LMRA to speak against unionization.
The ways courts have interpreted federal preemption it is dubious that any of these efforts will succeed. In
2002 the US District Court for California ruled against the state, interpreting employer free speech to include the
right to use state contract moneys to fight unionization even when state legislatures explicitly passed laws that said
they do not want taxpayer money spent in that way. Absent a Supreme Court decision reversing this and other
preemption decisions15, Congressional action would be needed to create space for states to operate in ways that they
have demonstrated they would like to do
But if Congress were to act, it would more likely allow states to undertake policies that could weaken as
well as strengthen unionism. Such a weakening of federal preemption could arguably gain the support of
representatives from states supportive of unionism and of representatives from states antithetical to unions. The
most plausible outcome would thus be increased diversity in state laws regulating private sector unionization,
475 U.S. 282 Argued December 9, 1985 Decided February 26, 1986
California Assembly Bill, AB1889, Cedillo, 00-872, State Funds: Unionization (2000).
The Supreme Court could respond to the California case by following the 1993 Rehnquist Court decision that
allowed Massachussetts to exclude nonunion contractors in the Boston Harbor construction case Building &
Construction Trades Council v. Associated Builders & Contractors of Massachusetts, 507 U.S. 218 (1993), known
as “Boston Harbor”, which would reduce the need for Congressional action to weaken federal preemption.
mimicking to some extent the diversity in public sector labor relations. Some states might introduce card checks for
union recognition. Others might try the quick elections that the Dunlop Commission recommended. Others might
enact place greater administrative regulatory burdens on unions, or discriminate against unionized firms in its state
contracts or increase the rights of employers to oppose unions. The answer to the title question hinges, then, on how
many workers might gain stronger rights to association and how many might see their rights weakened.
To answer the question, I compare estimates of the potential loss of coverage in states most likely to enact
state laws unfavorable to private sector collective bargaining with the potential gain in states most likely to enact
laws favorable to private sector unionism. My estimates are back-of-the envelope/excel computations that give
orders of magnitudes only. For my estimates I assume that all states with unfavorable public sector labor laws enact
equally unfavorable laws in the private sector (though they would presumably not be allowed to outlaw collective
bargaining). Column 1 of table 4 shows that in 2004, unions gave collective bargaining to 1.2 million persons in
these states. To see how many of these members might lose collective bargaining as a result of a change to a legal
regime less friendly to collective bargaining, I apply the elasticity of 0.5 from the regression of collective bargaining
in the public sector to favorable public sector laws given in table 2, line 4. My assumption is that a more unfavorable
environment would reduce coverage in the private sector by the same ln amount as favorable public sector laws
boosted coverage in the public sector. The calculation implies a loss of 0.5 million persons. With just 1.2 million
members in these states, the maximum loss is 1.2 million.
The second column of table 4 estimates the possible gain to unions from a more favorable private sector
legal environment in the states that have favorable public sector labor laws. Again, I use the elasticity of coverage
of 0.5 from table 2, line 4. The gain is 4 million persons – a far larger amount. The reason that on net collective
bargaining gains is that union membership is slight in the states that might enact legislation unfavorable to unions,
while it is non-negligible in states likely to enact favorable legislation.
Without specifying the specifics of laws and finding better ways to assess their impact (looking at the
Canadian changes? Examining the effect of prevailing wage laws on density in construction? using the within state
group variation, per Farber 2005), this suffices to make the point that if states were given authority for private sector
labor law, the net effect on collective bargaining would likely be positive.
5. Conclusion
Federal law has failed to deliver workers the representation and participation they want. The law leaves
millions of workers who want unions without workable ways to obtain union representation. It also restricts the
ability of firms to set up nonunion committees or councile for workers to meet and discuss issues with management
that many prefer to unions. The result is a huge unfilled demand for worker representation and participation in the
US, both union and nonunion. If the most recent Peter Hart surveys are on target, that unfilled demand for unionism
covers on the order of half of the nonunion work force in the country; while the unfilled demand for non-union
workplace groups approaches 80% of the US work force (Freeman and Rogers, chapter 1). Efforts to reform the
labor law have foundered. While there are no guarantees, turning the law regulating private sector labor relations
and/or its enforcement to the states cannot do much worse than the US labor law is doing now. Washington has
failed. Why not see if Sacramento and Bismarck, Albany and Oklahoma City, Des Moines and Detroit, Salt Lake
City and Madison, ... can do better?
Boston Federal Reserve New England Economic Review, June/July 1998
Brandeis, Louis, dissenting, New State Ice Co. v. Liebmann, 285 U.S. 262, at 311 (1932).
Cimini, Michael “1982-87 State and Local Government Work Stoppages and their Legal Background” Comensation
and Working Conditions, Fall 1998, pp 32-39
Clark, Jr. R. Theodore Public Sector Collective Bargaining: Past, Presetn, and Future
Donahue, John D. “Tiebout? Or Not Tiebout? The Market Metaphor and America's Devolution Debate” Journal of
Economic Perspectives, Volume 11, issue 4, 1997 pp. 73_81
Farber, Henry 2005 Union Membership in the United States: The Divergence between the Public and Private
September 2005
Farber, Henry S. 1988. \The Evolution of Public Sector Bargaining Laws." in When Public
Sector Employees Unionize, Richard B. Freeman and Casey Ichniowski, eds., Chicago.
University of Chicago Press. 129-166.
Ferejohn John and Barry R. Weingast, eds., The New Federalism: Can the States be Trusted? (Stanford: Hoover
Institution Press, 1997).
Fraser Institute Economic Freedom of North America 2005 Annual Report (by
Amela Karabegoviļ‚ƒ & Fred McMahon) http://www.freetheworld.com/efna.html
Freeman, Richard B. and Robert G. Valletta. 1988(b). The Effects of Public Sector Labor
Laws on Labor Market Institutions and Outcomes." in When Public Sector Employees
Unionize, Richard B. Freeman and Casey Ichniowski, eds., Chicago. University of Chicago Press. 81-106.
Freeman R and J. Rogers, “The Promise of Progressive Federalism,”in Making the Politics of Poverty and
Inequality, ed. Jacob Hacker, Suzanne Mettler, and Joseph Soss (New York, NY: Russell Sage Foundation,
Freeman R and J Rogers, What Workers Want, Cornell, 2 nd edition, forthcoming, 2006,
Gottesman, Michael H.“Rethinking Labor Law Preemption” Yale Journal on Regulation 7 (Summer 1990): 355-410
Greve, Michael ” Laboratories of Democracy Anatomy of a Metaphor
AEI Online (Washington) www.aei.org/publications/pubID.12743/pub_detail.asp.
Krueger, Carl, State Collective Bargaining Policies for Teachers http://www.ecs.org/clearinghouse/37/48/3748.pdf
Landau, Brent W. “State Employees and Soveriegn Immunity: Alternatives and Strategies for Enforcing Federal
Employment Laws” Harvard Journal on Legislation, vol 39, pp 169-212
Political Economy and Research Institute, U Mass, Decent Work in America The State-by-State Work Environment
Index, 2005 (by James Heintz, Jeannette Wicks-Lim, Robert Pollin)
Saltzman, Gregory M. 1988. \Public Sector Bargaining Laws Really Matter: Evidence from
Ohio and Illinois." in When Public Sector Employees Unionize, Richard B. Freeman
and Casey Ichniowski, eds., Chicago. University of Chicago Press. 41-80.
Web sites with data
Table 1: The Distribution of Groups of State and Local Employees Covered by Different
Public Sector Collective Bargaining Legislation, 1996
Type of Legislation
# of State Employee
Percentage of State Employee
Explicit Duty to Bargain
Implied Duty to Bargain
Authorized not required CB
Meet and Confer/present proposals
CB Barred
No Provision
Source: Tabulated from Kim Ruebens, update of Valletta-Freeman data set, downloadable from
NBER web site http://www.nber.org/publaw/.
* 27 States Favorable to CB: Alaska, California.Connecticut Delaware Florida, Hawaii, Illinois Iowa, Maine
Maryland, Massachusetts, Michigan, Minnesota, Montana, Nebraska,
New Hampshire, New Jersey, New Mexico, New York, Oregon, Ohio
Pennsylvania, Rhode Island, South Dakota, Vermont, Washington, Wisconsin
** 5 States Intermediate to CB: Arizona Indiana, Kansas, Missouri, W Virginia
*** 18 States Unfavorable to CB: Alabama , Arkansas ,Colorado, Georgia, Idaho , Kentucky Louisiana
,Mississippi, Nevada, North Carolina, North Dakota, Oklahoma, South Carolina Tennessee, Texas, Utah, Virginia,
Table 2: Statistical Analyses of Public Collective Bargaining Coverage, 2004
A. The level and range of Public Sector Coverage Compared to Private Sector Coverage
States with
States with
e laws
States with
48.6 (13.0)
28.0 (4.8)
22.2 (7.7)
22.9 to
23.0 to
10.4 to 39.6
Mean Density
9.4 (3.8)
8.9 (2.9)
5.1 (2.1)
Min, Max Density
3.6 to 16.7
4.7 to 11.6
2.2 to 11.0
Mean Difference
Min, Max Difference
18.1 to 56.5
15.9 to 22.3
7.5 to 29.9
Public Sector
Mean Density (SD)
Range of Density (Min,
Private Sector
Public Sector - Private
B. Coefficients and Standard Errors for the Regression of % Covered by Collective Bargaining in State on
Independent Variables, 2004
State CB LAW
Dependent variable
1 Percent Covered
20.6 (5.3)
2.Percent Covered
4.5 (4.1)
3.Ln % Covered
4. Ln % Covered
* in ln form in equations 3 and 4.
Source: Coverage, www.unionstats.com/
Laws, Kim Ruebens update of Valletta-Freeman data set http://www.nber.org/publaw/).
Table 3: Number of States with Relevant Private Sector Labor or Social Policy Regulation in Eight Different
Areas, by Public Sector Labor Relations Regulation
Number of Laws ( Probability of Having the Law)
All states
Minimum Wages Above Federal
Medicaid Eligibiity Criterion:
Earnings > 100 Percent
of Poverty
State Support for Postsecondary
Need-Based Aid 40% or more
of Federal Pell Grant Aid
Intermediate and
13 (48%)
11 (41%)
3 (13%)
10 (37%)
3 (13%)
23 (85%)
11 (41%)
10 (37%)
10 (43%)
21 (78%)
22 (96%)
Adopted Transitional Food Stamp
Option for Families Leaving
Prevailing Wage Law for State
State Earned
Income Tax Credit
OSHA Complete State Plan
Job Quality Standards attached to
Economic Development Incentives
Summary Statistic: Total number
of favorable laws (average per
Source: Laws, from Freeman and Rogers, 2006, Appendix.
Table 4: Number of private sector employees “at risk” of losing collective bargaining coverage in
states that would enact unfavorable CB state laws compared to the number of workers who might gain
coverage in states that would enact pro-CB state laws
States with public sector
labor laws unfavorable to
collective bargaining
States with public sector
labor laws favorable to
collective bargaining
Total Private Sector
31.5 million
63.9 million
Number of Unionized
1.2 million
6.3 million
Number of Nonunion
30.3 million
57.6 million
Potential Loss of Union
0.5 million to 1.2 million
Possible Gain in Union
4.1 million
Source: Numbers of employees by union status,
Estimated lost members in unfavorable law states, based on equation 4, table 2
Estimated gain in members in favorable law states, based on equation 4, table 2
The estimates assumed a change in the law would induce an ln increase/decrease in union membership of 0.5 ln