income gap and inequality
OdayGH
article2.pdf
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Labor Unions, Political Representation, and Economic Inequality John S. Ahlquist School of Global Policy and Strategy, University of California San Diego, La Jolla, California 92093; email: [email protected]
Annu. Rev. Polit. Sci. 2017. 20:409–32
First published online as a Review in Advance on March 9, 2017
The Annual Review of Political Science is online at polisci.annualreviews.org
https://doi.org/10.1146/annurev-polisci-051215- 023225
Copyright c© 2017 by Annual Reviews. All rights reserved
Keywords
class, redistribution, political economy, mobilization, civil society, partisanship, organizations
Abstract
Decades of research across several disciplines have produced substantial evi- dence that labor unions, on balance, reduce economic disparities. But unions are complicated, multifaceted organizations straddling markets and politics. Much of their equality-promoting influence occurs through their ability to reduce class-based inequity in politics and public policy. Declining union- ization across much of the developed world is eroding workers’ bargaining power. Reduced economic leverage puts pressure on union solidarity and weakens labor-based political movements. Important research design prob- lems and significant heterogeneity across unions, regions, countries, and time imply a continued need for more work.
409
Click here to view this article's online features:
• Download figures as PPT slides • Navigate linked references • Download citations • Explore related articles • Search keywords
ANNUAL REVIEWS Further
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
INTRODUCTION
“What do unions do?” was the question Freeman & Medoff (1984) posed in their landmark study. Their answer was that labor unions compress disparities. Exactly what disparities shrink, by how much and for whom, and at what cost and to whom have been the questions motivating the literature ever since. Increasing income inequality coincided with the erosion of union militancy and membership across the developed world, animating efforts to update Freeman & Medoff ’s conclusions (Rosenfeld 2014, Bennett & Kaufman 2007) and develop new research designs for better determining whether and how labor unions are causal agents.
In this review, I concentrate on what we know about the channels linking the declining for- tunes of labor unions and growing economic disparities. Three important findings emerge. First, as the earliest union organizers knew, unions’ roles in politics are at least as important as their industrial activities. Second, new data and longer time series indicate that past relationships be- tween unionism and the income distribution have weakened in the last 15–20 years. Third, the effects of unionism as well as the depth of union decline are profoundly conditioned by political institutions, which are themselves the result of past historical struggle and circumstance. This further challenges our ability to disentangle cause from effect, and we have yet to invent any durable organizational alternatives capable of mobilizing working-class citizens around economic concerns on the job, in the voting booth, and, sometimes, in the streets.
I begin with the large, interdisciplinary literature that attempts to estimate the effects of unions’ activities on predistribution—the pretax distribution of market incomes. Among rich democracies, there is robust evidence that unionization is associated with a more compressed wage distribution as well as reduced top income shares. Historically, the costs of compression in terms of unem- ployment or inflation depended on the organizational structure of the labor movement and the institutions governing wage bargaining and monetary policy. But changing technology and em- ployment imperatives appear to be generating bifurcated insider/outsider labor markets in both weakly and strongly unionized economies.
In rich democracies, unionization also correlates with greater redistribution of both income and economic risk, demonstrating that unions can be political actors. To the extent that unions tend to mobilize citizens who are further down the income distribution and less engaged in politics, unions can have equalizing effects on citizen engagement and, perhaps, the content of public policy. But there remain gaps, both theoretical and empirical, that prevent us from confidently asserting causality or bridging micro–macro gaps.
A lesson of this review is that unions are heterogeneous and evolving organizations (Marks 1989). Some are organized around workers sharing a common set of skills whereas others en- deavor to organize all workers in a firm or industry into a single body. Unions in the public sector face a very different organizing and bargaining environment from those in the private sec- tor. Internal union governance also varies; many are participatory mini-democracies, but some have been authoritarian hierarchies. Some unions maintain decentralized local organization while others operate at a removed level. Even within the same country and industry, unions’ scope of action varies. Some have long traditions of social movement–style activism in several domains while others hew closer to the stereotype of narrowly “economistic” interest groups. In some countries, unions have deep organizational ties to specific political parties, whereas in others, the union–party relationship is transactional and contingent. In some developing countries, nominal labor unions are more cogs in partisan, clientelist machines—or even tools of labor control and repression—than effective representatives of workers’ economic interests. This variation makes unions fascinating organizations for social scientists, but it also implies that macro-level variables, though clearly important, can obscure differences operating at the local, industrial, sectoral, or
410 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
regional levels. Newer research indicates that these differences can matter, especially for unions’ involvement in politics.
Unions and Their Differences
In a prototypical labor union, individuals independently come together in a formal organization to advocate for their interests as people who sell their labor for a wage. Unions generally represent shop-floor production workers rather than managers or professionals, but this need not be the case.
Unions are notoriously difficult to describe in terms of a single overarching objective because they must balance competing demands with uncertain trade-offs. Unions, like other organizations, can suffer agency problems between leaders and members. Nevertheless, scholars often proceed under the assumption that unions are broadly comparable across industries, countries, and time periods.
Unions in the public sector can be more controversial. In the public sector there is no profit to divide between workers and capital owners, nor are there direct competitive forces to restrain costs. Some argue that public sector union members, to the extent they mobilize members in elections, effectively select those with whom they will later have to bargain (Moe 2011). Public officials may therefore be less likely to resist union demands, making public sector organizing an attractive target as private sector unions become increasingly difficult to sustain. Public sector unions now make up large chunks of the labor movement in several countries, most notably the United States (Freeman 1988, Garrett & Way 1999). But it is not always obvious where to draw the line between public and private sector workers. Are workers in the state-owned airline or oil company private sector workers? Such questions have not been completely resolved; data and research on public sector unionism across countries remain limited.
The public/private distinction is even murkier for unions outside rich democracies. In such settings, especially authoritarian ones, the notion of independent unionism breaks down. In many Latin American countries, unions were historically part of an inclusionary corporatist structure underpinning import substitution development strategies. Union membership was compulsory, and governments appointed union leaders (Collier & Collier 1991). Developing Asia pursued a more exclusionary form of state-sanctioned unionism. In the former communist world, unions were state-sponsored, union leaders were Communist Party members, and union membership was nominally very high. Many of these so-called legacy unions have proven to be remarkably durable but, in some places, formerly government-controlled unions have succeeded in developing independent identities (Caraway 2012, Caraway et al. 2015). In others, workers’ organizations emerged parallel to pre-existing unions. We should be especially cautious when comparing union membership data in the developing world.
Inequality of What and for Whom?
This review emphasizes inequality in access to economic and political resources. On the economic side, existing scholarship typically considers wages and incomes rather than wealth, expenditures, or consumption. Most work on inequality as it relates to labor unions deals with differences across wage earners and households, but I briefly touch on work examining unionization’s consequences for between-group inequality, especially the gender wage gap and differences by race and national origin.
In any discussion of inequality—especially across countries—data availability, comparability, and quality are serious issues. To study inequality and unions we need high-quality data on wages and incomes at the worker or household levels, as well as matched information on union
www.annualreviews.org • Unions and Inequality 411
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
membership. Ideally these data are available longitudinally and linked to both individuals and firms. Such data are hard to assemble, even among rich democracies. They are largely absent in the developing world. The fact that union membership means something quite different in Poland and the Philippines compounds problems. For all these reasons, the literature reviewed here has a strong bias toward rich democracies, especially the United States.
When looking at political or representational inequality, we find less consensus about appro- priate measurement strategies. I highlight some of the recent work from American politics that attempts to link public opinion within different income groups to the voting behavior of legis- lators. But most work on unions in politics starts with the (defensible) assumption that political participation is positively correlated with income and socioeconomic status (SES). Any increase in the participation of the relatively poor or lower status therefore reduces political inequality, however defined.
UNIONS AND ECONOMIC INEQUALITY
Predistribution
Unions and the wage distribution in advanced countries. Unions aim to standardize, and ideally increase, wage rates across workers. Unionization can therefore be expected to reduce wage inequality among unionized workers while increasing wage disparities between the unionized and nonunionized (the union wage premium). But unionism’s effects on wages may spill over into the nonunion sector through several channels. Employers may raise wages to near-union levels to compete for labor or prevent unionization (the union threat effect); public policy may, to differing degrees, extend union wage agreements to workers who are not members (union coverage effects); union bargaining may restrain wages for managers and executives; and unions’ political activities may result in higher minimum wages, shorter work weeks, and norms of economic fairness (Western & Rosenfeld 2011).
Union wage premium. The notion of a union wage premium is hard to define in heavily unionized economies or places where wage bargaining institutions extend union-negotiated contracts well beyond the organized sector. An admittedly extreme example is France, with 2012 union density at approximately 8% but with near-universal union coverage; the United States in 2012 saw 11% density but only 12% coverage (Visser 2015). The literature on the union wage premium has therefore focused on Anglo-Saxon economies, especially the United States, where the gap between membership and coverage is small.
Since Freeman & Medoff ’s (1984) pioneering use of micro-level data, most studies of the union wage premium compare the wages of unionized and nonunionized workers after using regression to adjust for industry and location along with variables thought to predict unionization. The most recent contributions to this line of research return regression-adjusted private sector union wage premia on the order of 15–25% for men in the United States (Rosenfeld 2014). Consistent with Freeman & Medoff ’s earlier findings, there remains substantial heterogeneity in the union wage gap across industries and regions. The union wage premium is lower among female workers and in states or industries where product markets are more competitive or less unionized. The union wage premium in the public sector has been harder to determine, but existing evidence suggests that the wage benefit for public sector unionization in the United States is smaller than in the private sector.
In addition to wage premia, unionized workers in the United States are also much more likely to be offered and take advantage of fringe benefits such as health insurance and defined-benefit
412 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
pensions, adding to the union/nonunion gap in overall compensation (Buchmueller et al. 2002, Rosenfeld 2014). Efforts to establish a union safety premium by looking at workplace injuries have turned up contradictory and unstable results (Donado 2013).
Constructing a well-specified counterfactual for the union wage premium is a challenge (Card 2001, Card et al. 2004, DiNardo & Lee 2004), not least because employers under threat of union- ization might raise wages and benefits just enough to prevent their employees from joining up. In an influential paper, DiNardo & Lee (2004) take advantage of the American union recognition process, which often takes place via an employee vote, with the election administered by the Na- tional Labor Relations Board (NLRB). DiNardo & Lee use this voting procedure to produce a regression discontinuity (RD) design. Comparing manufacturing establishments where the union narrowly won recognition against establishments where the union narrowly lost, DiNardo & Lee find that unionization greatly increased the chances of a collectively bargained contract but showed no effect on average wages, employment, productivity, or establishment survival.
DiNardo & Lee’s findings run counter to both the older literature on unions’ effects and the fact that American employers go to great (and often illegal) lengths to prevent unionization (Bronfenbrenner 1994, Kleiner 2001). They note several limitations of their design that might help put this apparent contradiction in perspective. First, DiNardo & Lee’s findings are restricted to new unionization occurring during a period of union weakness and decline (1984–2000). Second, their indicator of unionization does not account for how much of an establishment’s workforce is actually covered by a particular union. Smaller or craft-based unions might represent only a small fraction of a firm’s workers whereas larger industrial unions may seek to represent all of them. Third, strong organizing drives may avoid the NLRB process altogether. NLRB elections near the win cut-off may involve unions that do not command enough bargaining power to effectively raise wages. Farber (2015) shows that, as unionization rates decline and organizing becomes more expensive, NLRB elections are less frequent, but those that do occur are more likely to be won by unions by greater margins.
DiNardo & Lee’s (2004) findings do not imply that unions are without consequences for the wage distribution within firms (Frandsen 2014). More problematic for the RD strategy, new research appears to show evidence of sorting (or, less generously, strategic interference in voting or counting) in the neighborhood of the election win threshold. Frandsen (2017) finds that unions are significantly more likely to lose close NLRB elections when the Board is controlled by Republican appointees, consistent with other work highlighting the politicization of the American union recognition process and the NLRB more generally (Moe 1985, Tope & Jacobs 2009). All this is cautionary for potential future applications of the NLRB election RD design to outcomes such as industrial accidents, workplace discrimination, or political activity.
Wage inequality. Arguably Freeman & Medoff ’s (1984) most important finding was that Amer- ican unions’ equalizing effects on the wage distribution outweighed the inequality-increasing union wage premium, leading to more compressed overall wage distributions in heavily unionized regions and industries. Rosenfeld (2014) and Western & Rosenfeld (2011) update those earlier findings and show that, once we condition on industry-region unionization, declining unionization accounts for approximately 30% of the 1973–2007 growth in private sector US wage inequality for men and approximately 20% for women. Card et al. (2004) find a similar association across the United States, United Kingdom, and Canada.
Brady et al. (2013) connect county-level working poverty rates in the United States with declin- ing state-level unionization. Important as this finding is, problems with the research design imply more work remains. For example, Autor et al. (2016) review recent work showing that trade shocks in the form of low-cost Chinese imports after China’s World Trade Organization accession have
www.annualreviews.org • Unions and Inequality 413
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
AUSAUSAUS AUSAUS
AUS AUSAUSAUSAUSAUS AUS
AUS AUS
AUSAUS AUS
AUS
AUS AUS AUSAUS
AUS AUS AUS
AUSAUS AUS
AUSAUS
AUSAUS AUS
AUTAUT AUT AUT AUTAUTAUT AUTAUTAUT
BELBELBELBEL BELBELBELBELBELBELBELBELBEL BELBEL
CANCANCANCAN CAN CAN CAN CANCANCANCANCAN
CAN CAN
CAN CANCAN
CHE CHECHECHE
CHECHECHECHE
CHE
CHL
CHL
CHL
CHL CHL
CHL
CZECZE CZECZECZECZE
CZE CZE CZECZECZECZECZECZECZE
CZECZECZE
DEU
DEU
DEU DEU
DEU DEU DEUDEU
DEU DEU DEUDEU
DEU DEU
DEU
DEU DEU
DEUDEUDEU
DEU
DEU
DNKDNKDNKDNKDNKDNK
ESP ESP
ESP
ESP ESP ESPESP ESP ESP
EST
EST
FINFINFINFINFIN FIN FIN
FIN FIN FINFINFINFINFINFINFINFINFIN
FIN FIN
FINFINFINFIN FINFIN
FIN FINFINFIN
FRAFRAFRAFRAFRAFRA FRAFRAFRA FRAFRAFRAFRAFRAFRAFRAFRAFRA
GBRGBRGBRGBR GBRGBRGBR
GBRGBRGBRGBR GBRGBRGBRGBR
GBRGBRGBRGBR GBRGBRGBRGBRGBRGBRGBRGBRGBR
GBR GBRGBRGBRGBRGBR
GRCGRCGRC
GRCGRC
GRC
GRC
HUN
HUN
HUN
HUN HUN HUN
HUN HUN
HUN
HUN IRL
IRL
IRL
IRL
IRL
IRL
IRL
IRLIRL IRL IRL
IRL
IRL
IRL
ISL ISL
ISL ISLISL
ISL
ISR
ISR
ITA
ITA
ITA ITAITA
ITA ITA
ITA
ITA
ITA ITA ITAITA
ITA
JPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPN JPN JPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPN
JPNJPN JPNJPNJPNJPNJPNJPNJPN
KORKOR KOR KOR
KOR KOR
KOR KOR KOR
KOR
KORKORKORKOR KORKOR
KOR KORKORKORKOR
KOR KOR
KORKOR KORKORKOR KOR
LTU
LTU
LUX
LVA
LVA
MEX
MEX MEX
MEX
MEX MEX
MEX
MEXMEX
NLD NLDNLD
NORNORNOR NOR NOR NORNORNORNOR NOR
NOR NORNORNORNOR NORNOR
NZLNZL
NZLNZLNZL NZL
NZL
NZL
NZLNZLNZL NZL
NZL NZL NZLNZLNZL NZLNZL NZLNZLNZLNZL NZLNZL
POL POL
POL POL
POL
POL
POLPOL POL
POLPOL POLPOL POLPOL
POL POL
POL POLPOL POL
POLPOL POL
POLPOL
PRT PRTPRT
PRT
PRTPRT
PRT PRT
PRT
SVKSVKSVK SVK
SVKSVK SVKSVKSVKSVKSVKSVK
SVN
SWE SWESWESWE SWE SWE SWESWESWESWE
SWE
SWE SWESWESWE SWESWE
SWESWESWE SWESWESWE
SWE SWE
SWE SWESWE
SWESWE SWESWESWE
USA USA USA USAUSAUSA
USAUSAUSAUSAUSAUSA USAUSA USAUSAUSAUSA
USAUSAUSAUSAUSAUSA USAUSAUSAUSA USA USAUSAUSA USA
USA
AUSAUS AUS AUSAUS AUS
AUSAUS AUSAUS AUS
AUSAUSAUS AUSAUS
AUSAUSAUS AUS
AUS AUS AUSAUSAUS
AUS AUSAUS
AUSAUS AUS AUSAUS
AUTAUTAUTAUTAUTAUTAUTAUTAUTAUT
BELBELBELBELBEL BELBEL BEL BEL BEL BEL BEL BELBELBELCAN
CANCANCANCANCANCANCANCANCANCANCANCANCANCANCAN CAN
CHE
CHECHECHE CHE CHECHECHE
CHE
CHL
CHL
CHL
CHLCHL
CHL
CZE CZE
CZECZE CZECZECZECZECZE
CZECZECZECZECZECZECZECZECZE
DEU DEU
DEU DEUDEUDEUDEU
DEUDEU DEUDEU DEU DEU DEU DEUDEUDEUDEUDEU
DEUDEUDEU
DNKDNK DNKDNKDNKDNK
ESPESP
ESP ESP
ESPESPESPESP
ESP
EST
EST
FIN FINFINFINFIN
FINFINFINFINFIN FINFINFIN FIN FIN
FINFINFINFINFINFIN FINFIN FINFINFINFINFINFINFIN
FRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRAFRA
GBR GBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBRGBR
GBRGBRGBRGBR GBRGBRGBRGBRGBRGBRGBRGBRGBR
GRC GRCGRC
GRC
GRC
GRC
GRC
HUN
HUN HUNHUN
HUN HUN
HUNHUN
HUN
HUN
IRL IRL
IRL
IRL IRL
IRLIRLIRLIRL
IRL
IRLIRLIRLIRL
ISLISL ISLISLISL
ISL
ISRISR
ITA
ITA
ITAITA
ITAITA ITA ITA
ITAITA
ITA ITA ITAITA
JPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPNJPN JPNJPNJPNJPNJPNJPNJPNJPNJPN
KOR
KOR
KOR
KORKOR KORKOR
KOR KORKORKORKOR
KORKORKORKORKORKORKORKORKORKORKOR KOR KOR KORKOR KOR KOR LTU LTU
LUX
LVA
LVA
MEXMEX
MEX
MEXMEX
MEX MEX
MEX
MEX
NLD
NLD NLD
NORNORNORNORNORNORNORNORNORNORNORNORNORNORNORNORNOR NZL
NZLNZL NZLNZLNZLNZL
NZL
NZL NZL NZLNZL
NZLNZL NZL NZLNZLNZL NZL NZL NZLNZL NZLNZLNZL
POLPOL POLPOL
POL
POLPOL POLPOL
POLPOLPOLPOL POLPOLPOLPOL POL
POL POLPOLPOL
POL POL
POL POL
PRT
PRTPRT PRT
PRT
PRT
PRT PRT PRT
SVK SVK
SVKSVK SVKSVKSVKSVKSVKSVKSVKSVK
SVN
SWESWESWE SWESWE SWESWESWESWESWESWE SWESWESWESWESWE
SWESWESWESWE SWESWESWESWESWESWESWESWESWESWE
SWE SWESWE
USAUSAUSA USA USAUSAUSA
USA USAUSAUSAUSAUSAUSAUSAUSAUSAUSAUSAUSAUSA
USAUSAUSAUSAUSAUSAUSA USAUSAUSAUSA USAUSA
1.25
1.50
1.75
2.00
2.25
2.50
25 50 75 100 Union density (%)
50 th
/1 0t
h pe
rc en
ti le
ra ti
o fo
r g ro
ss w
ag es
1980199020002010
Year
1.5
2.0
2.5
3.0
25 50 75 100 Union density (%)
90 th
/5 0t
h pe
rc en
ti le
ra ti
o fo
r g ro
ss w
ag es
a Upper half of wage distribution b Lower half of wage distribution
Figure 1 Unionization and inequality in (a) the upper and (b) lower halves of the gross wage distribution. All data points are country-years; lines are loess smoothers. Data sources: OECD (2016), Visser (2015).
important and long-lasting negative effects on local labor markets. These shocks could account for both union decline at the state level and rising working poverty at the county level, but this remains to be established.
Unionism is associated with a flatter overall wage distribution beyond the United States. Figure 1 plots union density against wage inequality for several rich and transition countries from 1980 to the present. Figure 1a displays the upper half of the wage distribution [the ratio of the 90th percentile (gross) wage to the median wage]; Figure 1b corresponds to the lower half (the ratio of the median to the 10th percentile wage). The relationship is clearly negative on a cross-national basis, with more compression in the bottom half of the wage distribution. The relationship appears to operate both across countries and within countries over time.
Several studies explore the relationships in Figure 1 using standard cross-national regressions (Wallerstein 1999, Bradley et al. 2001, Pontusson et al. 2002, Pontusson 2013). All look at rich democracies from the mid-1970s to the mid-1990s and find that union density is negatively associ- ated with wage inequality even after conditioning on a variety of covariates as well as country and time effects. These studies also turn up consistent negative associations between centralized wage setting and wage inequality. Pontusson et al. (2002) find that unionization’s effects are strongest in the bottom half of the income distribution; Pontusson (2013) updates this conclusion, reporting that the within-country relationship between union density and wage inequality has weakened since the mid-1990s.
414 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
But is unionism’s relationship to a more compressed wage distribution causal? This is harder to answer. The widening wage gaps in rich countries have been at least partly attributed to skill-biased technological change (SBTC) and the corresponding increase in relative demand for more highly educated workers. Acemoglu et al. (2001) develop a widely cited model explicitly linking SBTC to both deunionization and increased wage inequality. In their telling, unions are coalitions of workers with differing skill levels. SBTC puts upward pressure on skilled workers’ relative wages. This pressure not only widens wage gaps in the nonunion sector but also undermines the political coalition between skilled and unskilled union members. If skilled workers’ outside options become too good, they will leave unionized employment or push to decentralize bargaining, undermining the union wage compression effect.
Another strand of thinking links both deunionization and rising inequality with globalization. Depending on who tells the story, globalization takes the form of increased North–South trade; capital mobility abetting outsourcing and employer resistance (Slaughter 2007); or increased flows of migrants (Briggs 2001, Lee 2005). The immigration–deunionization link is tenuous and con- tested (Milkman 2006). The first wave of empirical work found that various forms of globalization played, at best, a bit part in the widening wage distribution of the 1980–1990s (Freeman 1995, Bradley et al. 2001, Scruggs & Lange 2002). But aforementioned work looking at the effect of US– China trade on local US labor markets gives reason to revisit these earlier conclusions. It seems that China’s entry into the global economy, especially since 2000, has affected wages. Whether this extends to workers’ ability to build and sustain unions is a question for future research.
Unions and wage inequality by race, national origin, and gender. Three groups—racial minori- ties, immigrants, and women—have faced intense employment discrimination and wage penal- ties. Unions, by standardizing pay and working conditions, can dramatically curtail management’s ability to arbitrarily favor particular workers. Unionization should therefore reduce wage gaps between disadvantaged groups and everyone else—if disadvantaged workers have access to unions and union jobs. Unions, employers, and governments have a mixed record on this score.
For decades, American unions largely excluded women and African Americans while cam- paigning for immigration restrictions. Foundational legislation protecting unionization excluded professions with high concentrations of black workers (Katznelson 2005, Frymer 2008). But af- ter World War II, American unions incorporated huge numbers of previously excluded workers. Rosenfeld & Kleykamp (2012), using Current Population Survey (CPS) data from the 1970s to the Great Recession, demonstrate that black workers, especially black women, were dispro- portionately represented in unionized jobs. This black/white unionization difference cannot be accounted for solely by positional variables (geography, industry, occupation), a finding Rosenfeld & Kleykamp interpret as evidence that disadvantaged workers actively sought the relative protec- tion from discrimination that unions provided. They further show that the decline of private sector unionism has had disastrous consequences for racial economic equality. Rosenfeld & Kleykamp generate model-based counterfactuals in which the black/white wage gap among women would be approximately 30% lower had unionization remained at its 1970s levels.
Immigration may exacerbate inequality to the extent that immigrants take jobs for lower wages than native workers do. Immigration may also put pressure on existing unions, since immigrants may be harder to organize owing to linguistic or cultural differences. For these reasons—along with simple prejudice—unions in immigrant-receiving countries, mainly Australia, Canada, and the United States, opposed immigration for several decades. Rosenfeld & Kleykamp (2009) use CPS data to look at the most recent wave of Hispanic immigration and find that Hispanics continue to join unions. They find that Hispanic unionization rates, unlike those for African Americans, can largely be explained by positional factors. Many American unions have recognized that organizing
www.annualreviews.org • Unions and Inequality 415
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
immigrants is crucial to their survival (Milkman 2006), but immigrants’ more precarious job status has made union gains harder to consolidate through the Great Recession (Catron 2013).
Gorodzeisky & Richards (2013), in one of the few comprehensive papers on unions and im- migrants in Europe, use four rounds of the European Social Survey (ESS) over 2002–2008 and find that noncitizens are less likely to be unionized across 14 rich European countries, but cross- country heterogeneity is stark. Countries where unions receive state support have more segmented labor markets and are particularly poor at incorporating migrants into the labor movement. Coun- tries operating under the Ghent system (in which unions administer unemployment insurance) see both higher levels of unionization and greater incorporation of migrants. Large-scale South– South economic migration is a relatively recent phenomenon, and we lack data on both migrants and their labor market experiences.
The situation for female workers is more complicated. The gendering of employment and the expectation that women would leave the labor force after marriage have long limited women’s access to unionized parts of the economy (Iversen & Rosenbluth 2011). In some countries union bargaining objectives, norms of fairness, and public policy were predicated on an assumed single- earner household. But standardized terms of employment and promotion along with an expanded public sector may attract more women into union jobs. The effect of unionization on wage inequal- ity between men and women is therefore ambiguous. Union density in rich democracies shows no association with the gap between median male and female wages. However, in the United States and United Kingdom, the gender wage gap narrowed at the same time unionization fell.
Unions and top incomes. Union-related pay compression extends well beyond blue-collar pro- duction workers. Union wage increases can reduce rents to management and capital owners. The process of negotiating with unions can make executive pay more visible and salient to other workers and the press. Higher executive pay may serve as a signal that the firm is doing well and capa- ble of paying higher wages, inducing unionized firms to restrain executive compensation growth. In some countries, workers are legally entitled to representation on corporate boards, enabling unions to have a direct say in executive pay. Union-controlled pension funds might likewise ex- ert pressure restraining executive compensation. Unions’ political activities may lead to higher income tax rates that discourage very high pay (Piketty et al. 2014).
Piketty (2014), in the most famous treatment of top income and wealth shares, attaches little explicit importance to unions. Nevertheless, early empirical work (DiNardo et al. 2000) using cross-sectional data sets in the United States turned up a negative relationship between union- ization and both the number of managers employed and the cash pay, i.e., excluding stock and other compensation, of managers and Chief Executive Officers (CEOs). Rosenfeld (2006b) uses CPS individual-level data to show that industry-region unionization is associated with reduced pay gaps between the median worker and median managers; this compression effect has grown weaker as unionization has declined. Shin (2014) connects industry-level unionization to lower relative pay to the top managers. Both studies rely on regional or industrial aggregations to as- sess unionization and, as such, are limited in what they tell us about unions in any particular firm. Gomez & Tzioumis (2013) have assembled the first panel of firm-level data on executive compensation and unionization in the United States. They find that unionization is associated with reduced compensation for top management on the order of 12%, coming in the form of reduced stock-based compensation in unionized firms. They also examine cross-firm executive pay and attempt to account for the possibility that unions might organize more profitable firms. Their instrument—union density of the state where the firms’ headquarters are located—could be questioned, but their basic finding that unions reduce executive compensation remains.
416 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
CEOs of American firms are consistently paid more than CEOs in other countries, both in absolute terms and relative to average worker pay (Murphy 1999). Fernandes et al. (2013) have produced the most detailed cross-national executive compensation data set to date. They continue to find that US-based executives in publicly listed firms receive more of their compensation packages in equity and are paid more than CEOs elsewhere, although their estimate of this gap is much smaller than previous findings. This data set has yet to be linked to unionization at either the national or firm level.
Unions’ actions can also have indirect consequences for higher incomes. Unionization and, especially, strikes can affect stock market values and hence executive pay and capital incomes (Abowd 1989, DiNardo & Hallock 2002, Lee & Mas 2012). As strikes have become increasingly rare, this effect is plausibly diminished (Rosenfeld 2006a, 2014). Agrawal (2012) and Ertimur et al. (2011) present evidence that AFL-CIO (American Federation of Labor and Congress of Industrial Organizations) affiliated union pension funds are particularly activist in opposing management positions in shareholder proxy votes, especially around compensation. Other indirect effects, such as whether unionized bargaining leads to greater media attention to executive pay, have gone largely unstudied.
In aggregate, union density is negatively correlated with top income shares cross-nationally (Scheve & Stasavage 2009, Huber et al. 2015) and in the United States (Volscho & Kelly 2012), even after conditioning on a variety of covariates. At the subnational level, however, the relationship differs. In the United States, where we have data on both union density (Hirsch & Macpherson 2003) and top income shares (Alvaredo et al. 2016) at the state level going back several decades, there is a strong longitudinal component; states in the 1960s had both higher unionization and lower top income shares than states in the 2000s. But across states unionization and top income shares are only weakly related, and the sign of this correlation actually switches over time; New York and Connecticut are among the most heavily unionized states, but they were also the most unequal by the late 1990s. Moller et al. (2009) turns up a similar finding from looking at county- level household income inequality.
The timing and magnitude of deunionization relative to growth in top income shares likewise varies. In some countries, such as Australia and Germany, deunionization preceded or coincided with growing top income shares. But American deunionization had been occurring more-or-less continuously for 25 years before top incomes began their steady climb. UK top income shares started increasing before unionization began to fall, and in the Netherlands a decrease in inequality preceded a decline in union density. In Denmark and France there has been very little change in top income shares despite declining union membership. Both the subnational experience in the United States and the wide variation across countries belies a simple causal connection between deunionization and growing inequality at the top. But the timing and magnitude differences across countries also make it unlikely that there is a simple overarching explanation (such as SBTC) that explains what we see in the data. National-level differences in institutions or policy are prime suspects.
Compression and the importance of institutions. Unionized firms pay relatively more for low- wage workers and, perhaps, a higher overall wage bill. We might then expect that unionization would reduce the employment of low-wage workers and perhaps lower employment overall. At the micro level, Freeman & Medoff (1984) argue that this need not be the case. To the extent unions reduce labor turnover, improve communication, reduce the employment of managers, and encourage training, they may be productivity enhancing. Data from the United States provides some evidence that unionized firms are less profitable and invest less in research and development
www.annualreviews.org • Unions and Inequality 417
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
while productivity differences are minimal (see Addison 2005 for a review). Again, there is variation in these estimates across industries, and their causal status is questionable.
At the macro level, two institutions—wage bargaining coordination and union contract extension—interact with unionization in determining how much and at what cost unionism com- presses the income distribution. If unions are organized and bargain at the enterprise level, then each union, representing a narrow slice of the workforce, fails to incorporate the costs of its wage settlement for other workers. High levels of unionization and decentralized bargaining can lead to higher unemployment and inflation along with extensive strike activity, presenting a collective action problem. In open economies, the wage pressure from unions in the nontradeable sector can impair the international competitiveness of the tradeable sector (Swenson 1991). If unions and firms were to bargain at the industry or sector level, then unions might better internalize the costs of their wage settlements, ideally leading to moderation, more egalitarian wage outcomes, and lower unemployment and inflation.
A significant theoretical and empirical literature investigates bargaining centralization and co- ordination. Notwithstanding some nuance, the literature generally finds that greater bargaining coordination is associated with both greater wage compression and better unemployment out- comes but that this relationship has weakened since the 1990s (see Driffill 2006 for a review). Inflation, on the other hand, is largely determined by the central bank’s monetary stance. Where an anti-inflation stance is credible, wage bargainers can factor this into their settlements ex ante (Iversen 1999).
The union coverage wedge—the difference between the union density rate and the proportion of workers covered by union-negotiated contracts—is also consequential. In some countries, no- tably France, Portugal, and Spain, the wedge is considerable, and union presence in the economy is unequally distributed across firms and sectors. In such situations union bargaining settlements, coordinated or otherwise, may represent the interests of only a few labor market insiders in larger firms who may fail to incorporate the costs of their settlements for workers in other situations. Extending these agreements across the economy may yield a more compressed wage distribution and limit growth in top income shares but at the cost of higher unemployment (Saint-Paul 2004, Aidt & Tzannatos 2008).
Attempts to establish causal relationships between wage bargaining institutions and the pretax income distribution are hamstrung by the paucity of cases and possible endogeneity both over time and across space. Bargaining coordination and wage moderation have distributional consequences within the labor movement and across firms (Wallerstein 1990). Engaging in coordinated bar- gaining requires unions and employer groups to have the organizational capacity to manage such conflict (Ahlquist 2010a, Baccaro & Simoni 2010). How does this come about? Taking a longer view, Iversen & Soskice (2009) and Thelen (2004) trace historical antecedents to the nineteenth century. Other work, both theoretical and empirical, shows that denser unionization and more co- ordinated wage bargaining emerged in places that were already more equal (Lee & Roemer 2005, Scheve & Stasavage 2009, Ahlquist 2010a, Beramendi & Rueda 2014). So although it seems clear that unionism’s effects on economic inequality are refracted through wage bargaining institutions, it also appears that these same institutions are themselves connected to the distribution of economic resources in earlier periods. Coordinated bargaining arrangements in several countries began un- raveling at around the same time income disparities began growing, raising the possibility of reverse causation. Acemoglu et al. (2012) present a model in which one country’s wage bargaining institutions and welfare state generosity are endogenous to choices made elsewhere. Whether the unequal American economy subsidizes Swedish egalitarianism remains to be established, but re- search designs that assume countries are conditionally independent are unlikely to resolve the issue.
418 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Unions and inequality in the rest of the world. It is not obvious that unions’ inequality- reducing effects in the Organization for Economic Cooperation and Development (OECD) should generalize more broadly. Traditional unions in developing countries tend to concentrate in formal- sector manufacturing and mining, the public sector, and state-owned enterprises, all of which tend to be bigger and pay better than agricultural, service, and informal enterprises. Unionization in such a context may actually exacerbate insider/outsider divisions in the labor market, dwarfing whatever equalizing role unions play in the firms and sectors where they are present. Existing research (reviewed by Freeman 2009) largely consists of country studies, sometimes relying on small, unrepresentative samples of workers and firms. Latin America, especially Mexico (Fairris 2003), has received the bulk of scholarly attention. Across studies, the union sector is frequently (but not always) associated with wage premia and lower wage dispersion compared to similar nonunion workers. Existing studies do not describe the relationship between unionization and the overall distribution of income.
Given the paucity of comparable longitudinal data on both unionization and the income dis- tribution for countries outside the OECD, scholars have turned to other indicators, most notably collective rights—the de jure extent of workers’ rights and violations in practice (Mosley & Uno 2007). Work in this vein turns up, above all, heterogeneity. Christensen & Wibbels (2014) find a positive relationship between inequality and more extensive collective labor rights among labor- abundant developing countries and the opposite among the labor scarce. Kerrissey (2015) turns up substantial regional heterogeneity using a different inequality data set and modeling strategy, pooling rich and developing countries. No work has yet been done looking at unionization or collective labor rights and top income shares in developing countries.
Redistribution
Unions have long been connected to redistributive politics and the rise of the modern welfare state. Studying redistribution requires data on both the market and disposable income distributions; earlier generations of scholars used imperfect proxies such as tax rates, government budgets, and labor market regulation. The Luxembourg Income Study harmonized micro-level data across countries, enabling the systematic study of redistribution, albeit for a relatively limited number of initial cases extending only as far back as the 1970s. Kenworthy & Pontusson (2005) show that redistribution, as measured by the difference between pre- and post-tax Ginis, has increased as pretax inequality has expanded almost everywhere. On a percentage reduction basis, however, more equal industrial democracies redistribute more.
First-generation economic theories of redistribution predict that democracies that are more unequal (prior to taxes and transfer payments) will redistribute more. The data clearly contradict this prediction, opening room for political agents that affect both the distribution of market income and the behavior of governments. The two main theoretical contenders here, power resources theory (Korpi 1983) and the class-compromise models (Iversen & Soskice 2006), have important political roles for unions. Bradley et al. (2001) and Iversen & Soskice (2006) take advantage of the Luxembourg Income Study data to develop cross-national regression studies of redistribution, finding that union density is positively associated with redistribution between the 1970s and late 1990s among OECD democracies. Pontusson (2013) extends the time frame up through 2010. Similar to his findings on wage inequality, Pontusson shows that the positive correlation between redistribution and unionization in the OECD has diminished since 1995. Building on arguments about insider/outsider politics (Rueda 2005, Iversen & Soskice 2015a), Pontusson conjectures that the changing composition of union movements has led to diminished solidarity in wage bargaining
www.annualreviews.org • Unions and Inequality 419
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
and a shift in unions’ political priorities away from broadly redistributive policies toward protecting labor market insiders.
UNIONS AND POLITICAL INEQUALITY
The relationship between unionization and redistribution presents a prima facie case that unions influence politics in ways that lead to reduced economic disparities. This is not a new realization. Labor unions have long been viewed as one of the few vehicles through which working- and lower- class citizens can wield some degree of political power (Lipset 1983). Unions figure prominently in major studies of American voter engagement and participation (Scholzman et al. 2012). As a simple descriptive matter, the rich and professional classes are vastly overrepresented among the American political elite (Carnes 2013). Union activities might equalize political representation in a number of ways.
Unions lobby legislators and marshal campaign contributions, activities generally reserved for the rich. Erikson (2015) highlights how differential rates of voting and political knowledge across income strata translate into differential rates of influence over policy, with lower-income citizens less influential. Unions provide intense social experiences that may transform workers into voters and more engaged citizens. Through hiring halls, team-based work, meetings, and social functions, unions enhance communication networks among members. Unions mount protests and strikes aimed at firms, governments, or even international organizations. They endorse candidates and policies and mount get-out-the-vote campaigns. Union members even tend to live in the same neighborhoods, especially when their work or dispatch sites are geographically concentrated (Ahlquist & Levi 2013).
Unions provide both job-based and political training. Members actively recruit one another into official union positions and electoral politics. In some countries, unions directly affiliate with political parties as part of a labor movement with reinforcing political and industrial segments. All these actions routinely pull workers into the political process, often in ways that systematically benefit parties of the Left and in support of redistributive policies. But there is also concern that de- clining unionization and changing union membership (older and more public sector workers) may reduce the power of unions’ mobilizing efforts and dilute the labor movement’s proredistributive message.
The academic literature on unions in politics, like other union-related scholarship, tends to emphasize rich democracies. The United States is particularly interesting because unionization is determined at a decentralized level. Once recognized, US unions do not compete among them- selves to represent the same workers in the same establishment, as happens in some European countries. American workers end up in a particular union for reasons that are almost entirely job-related, alleviating most concerns of self-selection and sorting based on political views.
Information, Beliefs, and Preferences
Might exposure to a union alter members’ political attitudes? How? One obvious channel is infor- mation. Some unions expend significant resources on outreach, education, and training. Iversen & Soskice (2015b) use cross-national survey data to show that union members are more politi- cally knowledgeable, engage in more political discussion, and tend to lean left. But beyond simply disseminating information, unions are important because they regularly call on members, who may only be there because of a job, to act jointly on a political issue. Being asked to contribute can reveal members’ latent or unexplored taste for political engagement, especially since many union members may come from backgrounds that do not encourage active political participation.
420 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Ahlquist & Levi (2013, p. 183) argue that “many individuals either have ill-formed preferences or have not been presented with the opportunity to act in ways consistent with their political com- mitments. Presenting members with an organized opportunity to act jointly and coherently forces them to solidify their preferences and presents them with information about their own efficacy.” Union engagement may help citizens learn not only about issues but also about themselves.
Labor histories are replete with unionists’ stories recounting the transformational effect their unions had not just on material situations but also on political views. Evidence beyond narrative accounts suggests that union exposure can alter members’ political attitudes. But there is hetero- geneity in both the strength of this effect and the domains in which it operates. Ahlquist & Levi (2013) and Ahlquist et al. (2014) look at unionization effects on voters’ attitudes toward interna- tional trade liberalization in one American union, the International Longshore and Warehouse Union (ILWU). Using matching techniques, Ahlquist et al. show that union exposure makes members less supportive of trade liberalization, especially the North American Free Trade Agree- ment, than otherwise similar nonmembers, even in a context where these workers have benefited from increased trade. Kim & Margalit (2017) find substantial heterogeneity in the union effect on trade policy preferences across industries and unions; unions that communicated more intensely with their members about trade show a bigger union effect. Looking specifically at the United Auto Workers (UAW), they show that when the UAW reversed its position on the Korea–US free trade deal, so too did the members; nonmembers who worked in the auto industry showed no such opinion change. Mosimann & Pontusson (2016) connect individual-level unionization with preferences for redistribution across several waves of the ESS. They find the biggest union effect among high-income union members.
Work to date shows that there is no one union effect on members’ political attitudes but rather various union effects. Different unions emphasize different policy areas and ask more or less of their members; these differences may operate at the branch or local level within the same union (Ahlquist & Levi 2013). Further research is needed to determine the stability and generalizability of these findings and weigh union effects against other competing demands on workers’ material and cognitive resources (Clark & Masters 2001).
Mobilization
Strikes and protest. Strikes and protest activity are unions’ most disruptive tactics. Industrial strikes can influence the income distribution directly, both by affecting firms’ profits and, possibly, raising wages (Rosenfeld 2006a). But as unionization has declined across the industrialized world, so too have strikes. In the United States, not only are strikes rare but the link between strikes and increased pay has disappeared as more strikes have occurred in the context of concessionary bargaining (Rosenfeld 2006a, 2014).
Employers are not the only audience for strikes. There is some evidence that large-scale po- litical strikes in OECD countries have increased in recent years (Hamann et al. 2012), especially around issues of employment protection, pensions, and welfare state retrenchment. Lindvall (2013) estimates a quadratic relationship between union density and political strikes, arguing that where workers are either very weak or very strong they cannot or need not engage in political strikes to show strength. Both Ahlquist & Levi (2013) and Kimeldorf (1988) document important differ- ences in political mobilization and orientation among unions in the same country, industry, and time period.
Union protest and strike activity are even more closely related to politics in the developing world, but in complicated ways. Strikes in post-Soviet Russia seem largely driven by elite conflict among regional governors and the central administration (Robertson 2007). In countries where
www.annualreviews.org • Unions and Inequality 421
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
unions maintain strong partisan connections, union strike and protest militancy depends on who is in power. Unions tend to be militant when their opponents are in power but quiescent in the face of reforms when their allies control the government (Teitelbaum 2010, Robertson 2004). Where partisan-allied unions must compete with one another for members, it becomes harder to moderate union behavior and coordinate on a political agenda. In such circumstances, high levels of union protest tend to emerge, especially during times of economic reform (Murillo 2001, Robertson 2004). But this protest may serve to undermine the union movement by showing politicians that the movement is divided and weak (Murillo 2001). Ultimately, the extent to which strikes affect firms’ decisions or government policy is a wide open area for research.
Political training. Unions routinely conduct meetings and ask members to vote for union offi- cers and contracts. Involvement in strikes and protest can transmit valuable political repertoires enabling participants to be more politically effective (McAdam et al. 2001). Union officers, drawn from the membership, often become important members of local communities and sometimes move into elected or appointed government positions. Sojourner (2013) finds that occupations with higher state-level unionization rates are represented with greater relative frequency among American state legislators. Carnes (2013) links declining unionization to a reduced number of American legislators with working-class backgrounds. Where unions are organizationally tied to political parties, becoming a union officer is a common pathway to a political career.
Research suggests that although unions can be “schools for democracy” (Terriquez 2011), this effect is not uniform across or even within unions. Unions with a more “social movement” ori- entation tend to have more democratic internal governance, invest more in developing members’ political and organizational skills, have ties to left-wing politics, and ask more of members than simple dues paying (Voss & Sherman 2000, Stepan-Norris & Zeitlin 2002). Terriquez (2011) documents how membership in a Los Angeles janitors’ union led immigrant parents to more critical involvement in their children’s schools. Importantly, however, this effect was limited to active members who worked in large buildings that were frequently contacted by union activists. Ahlquist & Levi (2013) show that members of dockworkers’ unions in two countries link their subsequent political activity to their early union education. Both these unions had long histories of participatory internal democracy, left-wing political commitments, and outreach efforts includ- ing frequent meetings, union newspapers, and even award-winning film units. Ahlquist & Levi turned up no such evidence of rank-and-file outreach or political training in the “business union” Teamsters or the corrupt, racket-ridden International Longshoremen’s Association.
Voter turnout. Unions manage to mobilize people not just in the streets but also in the voting booth, as demonstrated in a substantial literature documenting the union vote premium—the extent to which union members’ turnout rate exceeds that of nonmembers. Higher voter turnout has long been associated with more redistribution.
Radcliff & Davis (2000) show that higher levels of union density are associated with higher turnout, both cross-nationally and across American states. Leighley & Nagler (2007) analyze multiple waves of the American National Election Studies (ANES) and find that unionization boosts turnout most among lower-income groups in midterm elections. Flavin & Radcliff (2011) look at individual data cross-nationally using the International Social Survey Programme and find a consistent union vote premium as well as evidence for spillover to nonunion voters. Flavin (2017) shows that American states with higher levels of union membership show a closer congruence between the liberalism of lower-income survey respondents and the liberalism of state policy; no such relationship appears when Flavin looks at campaign contributions, suggesting that unions’ effects work largely through their ability to mobilize members.
422 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Rosenfeld (2010, 2014) looks at US national elections from 1984 to 2006 using CPS and ANES data. He makes two notable contributions. First, he breaks out union membership by sector. Noting that public sector workers (and union members) tend to be more educated than average, and the more educated are already more likely to vote, he identifies a significant difference between public and private sector union vote premia. He finds that unionized public sector workers turn out to vote in national elections at a rate approximately 2% higher than that of nonunion public sector workers once observable covariates have been incorporated. The same comparison in the private sector yields about a 7% union vote premium. The sector-based difference may help explain why the US union vote premium appears to have declined. It also implies that public sector unionism may not be as strong a force for equalizing political voice across income and class gradients.
Notwithstanding the important sectoral differences Rosenfeld identifies, he (and others) only considers voter turnout for US national elections. Moe (2011), in his study of US teachers’ unions, presents data showing that unionized teachers in Southern California are much more likely than the population average to turn out for off-cycle local elections, which tend to be low-information elections with very low overall turnout. Flavin & Hartney (2015) also look at teachers and highlight how policy can itself affect unionization. Using cross-state variation in the existence and timing of mandatory public sector collective bargaining rules, they find that teachers—even nonunion members—in states with mandatory collective bargaining are significantly more politically active. Aniza (2011) and Berry & Gersen (2011) look at the effects of off- versus on-cycle elections for outcomes such as teacher pay. Both papers find a pay premium of approximately 3% for teachers where school board elections were held off cycle. So while it may be that public sector union engagement has minimal effect on members’ turnout in national elections, consequences for lower-level government could still be substantial in decentralized systems such as the United States.
Rosenfeld (2014) echoes Leighley & Nagler’s (2007) finding that union vote premia vary across the population. But instead of looking at income, Rosenfeld compares private sector union vote premia across education strata. He finds that the union vote premium is largest (approximately 11%) among those without a high school diploma, declining to 4% among those with a college degree. Kerrissey & Schofer (2013), although not distinguishing between public and private sec- tors, find a similar pattern: The strongest association between union membership and political participation is found among lower-SES individuals. Private sector unionism is strongly associated with increased participation among those least likely to participate otherwise.
Bryson et al. (2014) uses the ESS to document union vote premia across 29 countries. By exam- ining how estimated union vote premia change after conditioning on income, they also interrogate the three possible channels through which the union effect may flow. Union membership remains a strong predictor of voter turnout even after conditioning on income, but the vote premium is smaller, implying that some of the union turnout effect works through increasing members’ incomes. Bryson et al. also find that former union members continue to vote and participate in politics at higher rates than “never members,” consistent with habituation or socialization in the union. But this premium is smaller than for current union members, implying that social pressure within the union is also a channel. The heterogeneity across unions and institutional contexts means there is plenty of room for additional work here.
There are two notable weaknesses in existing work on unions and the union vote premium. First, although we have robust regression-controlled comparisons yielding consistent estimates in the neighborhood of 5–10%, the literature has not yet settled on a clearly defined counterfactual scenario of interest. Some possibilities include:
1. What is the effect on an individual’s decision to vote if the establishment where she works becomes unionized?
www.annualreviews.org • Unions and Inequality 423
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
2. What is the effect on the turnout decision of one member of a household if another household member takes a union job?
3. What is the effect of an additional year as a union member on an individual’s decision to vote?
4. What is the effect on overall turnout of a change in the level of unionization? Each is an important question with different underlying causal mechanisms highlighting the
fact that unions’ political activities likely affect the voting decisions of both union members and nonmembers. Research design challenges are nontrivial. For example, at the macro level, declining unionization is associated with greater economic inequality, which is, in turn, linked to reduced turnout in the lower half of the income distribution (Iversen & Soskice 2015b). But this relationship appears conditional on electoral institutions and the structure of the party system (Anderson & Beramendi 2012), which may itself be a (complicated) function of historical levels of union strength (Huber & Stephens 2001).
Second, all the work linking union membership and turnout relies on survey respondent self- reports. It is well known that more educated, higher-income, more partisan, and more politically engaged survey respondents are more likely to overreport having voted, which can lead to erro- neous descriptions of the electorate (Karp & Brockington 2005, Ansolabehere & Hersh 2012). Given findings connecting union membership to political knowledge, engagement, and norms of participation, it seems plausible that the reported union turnout effect could be overstated, per- haps significantly so. Resolution of this issue awaits further research and micro-level data linking validated voting to union membership (ideally, also validated).
Parties and Policies
Unions almost always have some sort of alliance with political parties. These linkages vary and interact with electoral institutions in ways that can affect both the content and implementation of public policy. Where the extension of the franchise to the (male) working class predated in- dustrialization, then unions generally developed independently of political parties (Lipset 1983). Before the Great Depression, the American labor movement maintained a studied distance from partisan politics. But the union-enabling legislation in the New Deal laid the foundation for an important alliance between many unions and the Democratic Party. Union voters still tend to skew Democratic (Rosenfeld 2014). But union–party linkages in the United States have mostly taken the form of lobbying, campaign contributions, and partisan voter turnout efforts (Burns et al. 2000). Ever since the Taft-Hartley Act, American unions have been more successful in lobbying for broad-based labor market policy than in securing legislation directly beneficial to unions and union organizing (Freeman & Medoff 1984). As unionization rates have declined, unions’ legisla- tive success rate at the federal level has also waned, culminating in the stinging 2009 defeat of the Employee Free Choice Act despite Democratic control of the House, the presidency, and a supermajority in the Senate.
Where movements for democracy happened alongside or after industrialization, we see parties and unions intimately connected, with one providing funding for the other and often sharing offices, officers, and coordinated industrial and political strategy. In some countries, especially those operating under mixed and proportional electoral systems with decentralized bargaining, this has led to fragmented union movements with different, competing unions affiliated with different political parties. In other countries, notably Mexico, government-sponsored unions are integral parts of clientelist mobilization networks (Larreguy et al. 2014).
The self-reinforcing strength of the political left and organized labor is of long-standing interest for students of the welfare state. Power resources theory holds that working-class
424 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
movements, primarily unions, mobilized voters into Labor and Social Democratic parties that then implemented industrial relations policies benefiting unions, as well as redistributive and social insurance policies benefiting those in the bottom half of the income distribution. Other models emphasize the importance of electoral institutions in generating leftist government (Iversen & Soskice 2006). Pontusson & Rueda (2010) use a series of cross-national regressions to examine Left party platforms in ten developed democracies from the late 1960s to the turn of the twentieth century. They find not only that more heavily unionized countries have more leftist Left parties but also that voter turnout—itself linked to unionization—interacts with growing inequality to pull the Left further in that direction. Declining unionization and parties’ willingness to adopt pro-worker platforms are therefore linked directly (via unions’ effects on the parties), indirectly (via unions’ ability to mobilize voters in the bottom half of the distribution), and tangentially (via unions’ effects on the pretax income distribution).
Unions are sometimes directly integrated into policy making. The most developed instances are the bi-and tripartite policy bodies found in the highly unionized and coordinated northern European countries. Policy making is explicitly consultative across a number of specific though limited domains, including employment protections and labor market policies, taxation, pensions, and training. In some instances, union federations must assent to policy changes.
In the absence of such bodies, unions have still been party to major reform initiatives surround- ing macroeconomic crisis adjustment (Avdagic et al. 2011). The most visible such events are social pacts: formal, written, time-bound, and publicly announced agreements between unions, govern- ments, and possibly employers, in which “labor unions pull their punches in wage negotiation or assent to changes in labor market regulations in exchange for social spending, taxation, or other policies they prefer” (Ahlquist 2010b, p. 573). Pacts are closely linked to the electoral needs of center-Left parties and unions’ abilities to credibly commit to working with government. For ex- ample, in the early 1980s, both Australia and New Zealand struggled to deal with major structural and balance-of-payments crises. Ahlquist (2011) shows that long-term connections between the Labo(u)r parties and the major union associations explain why a pact was sustainable in Australia but not in New Zealand. Ultimately, adjustment costs were shared more equitably in Australia; growth in Australian income inequality was relatively muted.
Beyond their role as public sector employers, governments are sometimes involved in wage set- ting. The most obvious cases are minimum wage laws, which have been linked to lower inequality (Lee 1999, Autor & Smith 2016). Unions have frequently—but not always—supported minimum wages. The German trade union movement long opposed a government-set minimum wage, preferring to bargain and set wage floors in each sector. But declining union density, increased use of temporary employment contracts, and increased immigration limited the extent to which collectively bargained wages reached the bottom of the distribution. The major German unions changed their position decisively in 2011, and Germany introduced its first federal minimum wage in 2015.
Unions and other worker representation bodies can also play roles in implementing or enforcing inequality-reducing policies. Union presence seems to improve the enforcement of minimum wage laws and other labor market standards in the United States (Weil 1999, Fine & Gordon 2010). In the developing world, the presence of independent worker organizations appears to improve labor standards in parts of international supply chains (Berliner et al. 2015).
Democracy
What about democracy? Important theories of democratization put economic distributive conflict and inequality front-and-center; democracy is generally viewed as more redistributive and better
www.annualreviews.org • Unions and Inequality 425
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
for poor and working-class citizens than authoritarian alternatives. In the West, unions have been associated with the long march of democratization over the nineteenth and twentieth centuries (Collier 1999). South African unions and their federation played important roles in the transi- tion from Apartheid (Wood 2000). Unions can be central to civil society networks pushing for democratic opening and better governance (Lee 2007).
But strikes and union alliances are almost never sufficient to induce a regime transition on their own. Unions, even if successful at mobilizing workers under authoritarian systems or as voters, are not always prodemocratic elements (Valenzuela 1989, Levitsky & Mainwaring 2006). Unions deeply incorporated into populist or Marxist parties can end up inhibiting democratization, even when independent labor organizations are pushing in the opposite direction (Levitsky 2001). Union leaders, when insulated from rank-and-file pressure, can become co-opted by parties or even criminal elements. Whether unions are part of pro- or antidemocratic coalitions can vary across cases and across unions within a country, depending on the instrumental benefits offered to union leaders and members as well as the expected outcomes under different regime types.
Unions are demonstrably political organizations. On balance, unions have significant equalizing effects both in terms of political representation and policy output. But the manner and extent of unions’ political integration and their success at influencing policy varies substantially over time and across countries. Unsurprisingly, the strength of unions’ political influence appears to be waning as union membership declines. Important work remains to determine whether unions’ political resources are shifting toward efforts defending the privileges of labor market insiders at the expense of broader redistributive concerns.
CONCLUSION
Labor unions are associated with more compressed distributions of wages and incomes, both before and after government taxes and spending. In their industrial and collective bargaining activities, unions have been able to raise wages and compress the overall income distribution, including slowing the relative growth in top incomes. But wage bargaining institutions condition how effective these efforts might be. In the face of changing technology and competition from low-wage parts of the world, it appears unions’ strong compression effects on pretax incomes may be waning. The erosion of unionization and the changing composition of union membership is putting pressure on bargaining solidarity as well as unions’ ability to deliver outcomes beneficial to low-paid workers and labor market outsiders.
Much of the connection between unions and the distribution of economic resources occurs through political channels. There is extensive evidence, both macro and micro, linking union membership to increased political knowledge, interest, and turnout as well as other forms of par- ticipation. Union members vote systematically to the Left of similar nonunion counterparts. This is not just a matter of self-selection; union engagement appears to change some members’ beliefs about both politics and personal political efficacy. Unions’ alliances with political parties have also provided avenues into the policy-making process, whether through candidate recruitment, party platforms, lobbying and contributions, or explicit policy bargains. In authoritarian contexts, independent unions become inherently political, since direct collective bargaining and striking also challenge state control. With varying levels of effectiveness and durability, unions’ political activities appear to generate policies tending to equalize the distribution of income.
There is tremendous heterogeneity across individual unions, sectors, states, countries, and time periods. Unions vary in organizational attributes and institutional environments in ways that affect their bargaining and strike behavior as well as political stances. They vary in the weight they place on political activity relative to industrial relations and the extent to which they engage and make
426 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
demands of their members. Sectoral differences affect both member demographics and unions’ tools and incentives as they make demands.
Unionization is part of a complicated web of associations and plausible causal relationships operating endogenously and dynamically. Union activity influences parties and policy, but policy and institutions also affect unionization rates, the forms unions take, and the degree to which they affect the income distribution. External shocks and structural changes are important, but they too are refracted through existing institutions. The policy choices and economic outcomes in one country have consequences for those in others. And, of course, history matters.
Among OECD democracies, we have too many explanations chasing too few data points that are themselves interdependent in both time and space. Perhaps we will discover a convincing instrument for unionization. More likely is that research designs explicitly taking advantage of heterogeneity in context and population will incrementally provide pieces to the puzzle. Such designs might take advantage of large-scale shocks (such as British exit from the European Union, should it come to pass) or more micro events (such as deaths of union leaders or heads of state). Crucial to improved understanding will be our ability to build and sustain better administrative data, especially in the United States and many developing countries.
DISCLOSURE STATEMENT
The author is not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
In writing this article, I benefited from the hospitality of Nuffield College at the University of Oxford and at the Center for Advanced Study in the Behavioral Sciences at Stanford University. I thank Mitch Downey, Thomas Flaherty, Peter Gourevitch, Lane Kenworthy, and the editors of the Annual Review of Political Science for helpful comments. John Seungmin Kuk provided research assistance.
LITERATURE CITED
Abowd JM. 1989. The effect of wage bargains on the stock market value of the firm. Am. Econ. Rev. 79:774–800 Acemoglu D, Aghion P, Violante GL. 2001. Deunionization, technical change and inequality. In Carnegie-
Rochester Conference Series on Public Policy, 55:229–64. Amsterdam: Elsevier Acemoglu D, Robinson JA, Verdier T. 2012. Can’t we all be more like Scandinavians? Asymmetric growth and
institutions in an interdependent world. NBER Work. Pap. No. 18441. Natl. Bur. Econ. Res., Cambridge, MA
Addison JT. 2005. The determinants of firm performance: unions, works councils, and employee involvement/high-performance work practices. Scottish J. Polit. Econ. 52:406–50
Agrawal AK. 2012. Corporate governance objectives of labor union shareholders: evidence from proxy voting. Rev. Financ. Stud. 25:187–226
Ahlquist JS. 2010a. Building strategic capacity: the political underpinnings of coordinated wage bargaining. Am. Polit. Sci. Rev. 104:171–88
Ahlquist JS. 2010b. Policy by contract: electoral cycles, parties and social pacts, 1974–2000. J. Polit. 72:572–87 Ahlquist JS. 2011. Navigating institutional change: the accord, rogernomics, and the politics of adjustment in
Australia and New Zealand. Comp. Polit. Stud. 44:127–55 Ahlquist JS, Clayton AB, Levi M. 2014. Provoking preferences: unionization, trade policy, and the ILWU
puzzle. Int. Organ. 68:33–75
www.annualreviews.org • Unions and Inequality 427
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Ahlquist JS, Levi M. 2013. In the Interest of Others: Organizations and Social Activism. Princeton, NJ: Princeton Univ. Press
Aidt TS, Tzannatos Z. 2008. Trade unions, collective bargaining and macroeconomic performance: a review. Ind. Relat. J. 39:258–95
Alvaredo F, Atkinson AB, Piketty T, Saez E, Zucman G. 2016. WID—the world wealth and income database. http://wid.world. Accessed May 27, 2016
Anderson CJ, Beramendi P. 2012. Left parties, poor voters, and electoral participation in advanced industrial societies. Comp. Polit. Stud. 45:714–46
Aniza SF. 2011. Election timing and the electoral influence of interest groups. J. Polit. 73:412–27 Ansolabehere S, Hersh E. 2012. Validation: what big data reveal about survey misreporting and the real
electorate. Polit. Anal. 20:437–59 Autor DH, Manning A, Smith CL. 2016. The contribution of the minimum wage to US wage inequality over
three decades: a reassessment. Am. Econ. J. Appl. Econ. 8:58–99 Autor DH, Dorn D, Hanson GH. 2016. The China shock: learning from labor-market adjustment to large
changes in trade. Annu. Rev. Econ. 8:205–40 Avdagic S, Rhodes M, Visser J, eds. 2011. Social Pacts in Europe: Emergence, Evolution, and Institutionalization.
Oxford, UK: Oxford Univ. Press Baccaro L, Simoni M. 2010. Organizational determinants of wage moderation. World Polit. 62:594–635 Bennett JT, Kaufman BE, eds. 2007. What Do Unions Do? A Twenty-Year Perspective. New York: Transaction Beramendi P, Rueda D. 2014. Inequality and institutions: the case of economic coordination. Annu. Rev. Polit.
Sci. 17:251–71 Berliner D, Greenleaf A, Lake M, Levi M, Noveck J. 2015. Labor Standards in International Supply Chains:
Aligning Rights and Incentives. London: Edward Elgar Berry CR, Gersen JE. 2011. Election timing and public policy. Q. J. Polit. Sci. 6:103–35 Bradley D, Huber E, Nielsen SMF, Stephens JD. 2001. Distribution and redistribution in postindustrial
democracies. World Polit. 55:193–228 Brady D, Baker RS, Finnigan R. 2013. When unionization disappears: state-level unionization and working
poverty in the United States. Am. Sociol. Rev. 78:872–96 Briggs V. 2001. Immigration and American Unionism. Ithaca, NY: Cornell Univ. Press Bronfenbrenner K. 1994. Employer behavior in certification elections and first-contract campaigns: implica-
tions for labor law reform. In Restoring the Promise of American Labor Law, ed. S Friedman, R Hurd, R Oswald, R Seeber, pp. 75–89. Ithaca, NY: ILR Press
Bryson A, Gomez R, Kretschmer T, Willman P. 2014. What accounts for the union member advantage in voter turnout? Evidence from the European Union, 2002–2008. Relations Industrielles/Ind. Relat. 69:732–65
Buchmueller TC, DiNardo J, Valletta RG. 2002. Union effects on health insurance provision and coverage in the United States. Ind. Labor Relat. Rev. 55:610–27
Burns PF, Francia PL, Herrnson PS. 2000. Labor at work: union campaign activities and legislative payoffs in the US House of Representatives. Soc. Sci. Q. 81(2):507–22
Caraway TL. 2012. Pathways of dominance and displacement: the varying fates of legacy unions in new democracies. World Polit. 64:278–305
Caraway TL, Cook ML, Crowley S, eds. 2015. Working Through the Past: Labor and Authoritarian Legacies in Comparative Perspective. Ithaca, NY: Cornell Univ. Press
Card D. 2001. The effect of unions on wage inequality in the US labor market. Ind. Labor Relat. Rev. 54:296–315 Card D, Lemieux T, Riddell WC. 2004. Unions and wage inequality. J. Labor Res. 25:519–59 Carnes N. 2013. White-Collar Government. Chicago: Univ. Chicago Press Catron P. 2013. Immigrant unionization through the great recession. Am. Sociol. Rev. 78:315–32 Christensen D, Wibbels E. 2014. Labor standards, labor endowments, and the evolution of inequality. Int.
Stud. Q. 58:362–79 Clark P, Masters MF. 2001. Competing interest groups and union members’ voting. Soc. Sci. Q. 82:105–16 Collier RB. 1999. Paths Toward Democracy: The Working Class and Elites in Western Europe and South America.
New York: Cambridge Univ. Press Collier RB, Collier D. 1991. Shaping the Political Arena. Princeton, NJ: Princeton Univ. Press
428 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
DiNardo JE, Hallock KF. 2002. When unions “mattered”: the impact of strikes on financial markets, 1925– 1937. Ind. Labor Relat. Rev. 55:219–33
DiNardo JE, Hallock KF, Pischke JS. 2000. Unions and the labor market for managers. Disc. Pap. 2418, Cent. Econ. Policy Res., London
DiNardo JE, Lee DS. 2004. Economic impacts of new unionization on private sector employers: 1984–2001. Q. J. Econ. 119:1383–441
Donado A. 2013. Why do unionized workers have more occupational injuries? Disc. Pap. No. 551, Univ. Heidelberg, Germany
Driffill J. 2006. The centralization of wage bargaining revisited: What have we learnt? J. Common Market Stud. 44:731–56
Erikson RS. 2015. Income inequality and policy responsiveness. Annu. Rev. Polit. Sci. 18:11–29 Ertimur Y, Ferri F, Muslu V. 2011. Shareholder activism and CEO pay. Rev. Financ. Stud. 24:535–92 Fairris D. 2003. Unions and wage inequality in Mexico. Ind. Labor Relat. Rev. 56:481–97 Farber H. 2015. Union organizing decisions in a deteriorating environment: the composition of representation
elections and the decline in turnout. ILR Rev. 68:1126–56 Fernandes N, Ferreira MA, Matos P, Murphy KJ. 2013. Are US CEOs paid more? New international evidence.
Rev. Financ. Stud. 26(2):323–67 Fine J, Gordon J. 2010. Strengthening labor standards enforcement through partnerships with workers’ or-
ganizations. Polit. Soc. 38:552–85 Flavin P. 2017. Labor union strength and the equality of political representation. Br. J. Polit. Sci. In press. doi:
10.1017/S0007123416000302 Flavin P, Hartney MT. 2015. When government subsidizes its own: collective bargaining laws as agents of
political mobilization. Am. J. Polit. Sci. 59:896–911 Flavin P, Radcliff B. 2011. Labor union membership and voting across nations. Elect. Stud. 30:633–41 Frandsen BR. 2014. The surprising impacts of unionization on establishments: evidence from matched employer-
employee data. Work. pap., Brigham Young Univ. Frandsen BR. 2017. Party bias in union representation elections: testing for manipulation in the regression
discontinuity design when the running variable is discrete. Adv. Econometrics 38:In press Freeman RB. 1988. Contraction and expansion: the divergence of private sector and public sector unionism
in the United States. J. Econ. Perspect. 2:63–88 Freeman RB. 1995. Are your wages set in Beijing? J. Econ. Perspect. 9:15–32 Freeman RB. 2009. Labor regulations, unions, and social protection in developing countries: market distortions or
efficient institutions? NBER Work. Pap. No. 14789, Natl. Bur. Econ. Res., Cambridge, MA Freeman RB, Medoff JL. 1984. What Do Unions Do? New York: Basic Books Frymer P. 2008. Black and Blue: African-Americans, the Labor Movement, and the Decline of the Democratic Party.
Princeton, NJ: Princeton Univ. Press Garrett G, Way C. 1999. Public sector unions, corporatism, and macroeconomic performance. Comp. Polit.
Stud. 32:411–34 Gomez R, Tzioumis K. 2013. Unions and executive compensation. Disc. Pap. 720, Cent. Econ. Perform., London Gorodzeisky A, Richards A. 2013. Trade unions and migrant workers in Western Europe. Eur. J. Ind. Relat.
19:239–54 Hamann K, Johnston A, Kelly J. 2012. Unions against governments: explaining general strikes in Western
Europe, 1980–2006. Comp. Polit. Stud. 46(9):1030–57 Hirsch BT, Macpherson DA. 2003. Union membership and coverage database from the current population
survey. Ind. Labor Relat. Rev. 56:349–54 Huber E, Huo J, Stephens JD. 2015. Power, markets, and top income shares. Work. Pap. No. 404, Kellogg Inst.
Int. Stud., Notre Dame, IN Huber E, Stephens JD. 2001. Development and Crisis of the Welfare State: Parties and Policies in Global Markets.
Chicago: Univ. Chicago Press Iversen T. 1999. Contested Economic Institutions. New York: Cambridge Univ. Press Iversen T, Rosenbluth F. 2011. Women, Work, and Politics: The Political Economy of Gender Inequality. New
Haven, CT: Yale Univ. Press
www.annualreviews.org • Unions and Inequality 429
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Iversen T, Soskice D. 2006. Electoral institutions and the politics of coalitions: why some democracies redis- tribute more than others. Am. Polit. Sci. Rev. 2:100(2):165–81
Iversen T, Soskice D. 2009. Distribution and redistribution: the shadow from the nineteenth century. World Polit. 61:438–86
Iversen T, Soskice D. 2015a. Democratic limits to redistribution: inclusionary versus exclusionary coalitions in the knowledge economy. World Polit. 67:185–225
Iversen T, Soskice D. 2015b. Information, inequality, and mass polarization: ideology in advanced democracies. Comp. Polit. Stud. 48:1781–813
Karp JA, Brockington D. 2005. Social desirability and response validity: a comparative analysis of overreporting voter turnout in five countries. J. Polit. 67:825–40
Katznelson I. 2005. When Affirmative Action Was White. New York: W.W. Norton Kenworthy L, Pontusson J. 2005. Rising inequality and the politics of redistribution in affluent countries.
Perspect. Polit. 3(3):449–71 Kerrissey J. 2015. Collective labor rights and income inequality. Am. Sociol. Rev. 80(3):626–53 Kerrissey J, Schofer E. 2013. Union membership and political participation in the United States. Soc. Forces
91:895–928 Kim SE, Margalit Y. 2017. Informed preferences? The impact of unions on workers’ policy views. Am. J. Polit.
Sci. In press. doi: 10.1111/ajps.12280 Kimeldorf H. 1988. Reds or Rackets? The Making of Radical and Conservative Unions on the Waterfront. Berkeley:
Univ. Calif. Press Kleiner M. 2001. Intensity of management resistance: understanding the decline of unionization in the private
sector. J. Labor Res. 22:519–40 Korpi W. 1983. The Democratic Class Struggle. London: Routledge Larreguy H, Olea C, Querubin P. 2014. The role of labor unions as political machines: evidence from the case of the
Mexican Teachers’ Union. Work. pap., Harvard Univ. Lee CS. 2005. International migration, deindustrialization and union decline in 16 affluent OECD countries,
1962–1997. Soc. Forces 84:71–88 Lee CS. 2007. Labor unions and good governance: a cross-national, comparative analysis. Am. Sociol. Rev.
72:585–609 Lee DS. 1999. Wage inequality in the United States during the 1980s: rising dispersion or falling minimum
wage? Q. J. Econ. 114:997–1023 Lee DS, Mas A. 2012. Long-run impacts of unions on firms: new evidence from financial markets, 1961–1999.
Q. J. Econ. 127(1):333–78 Lee W, Roemer JE. 2005. The rise and fall of unionised labour markets: a political economy approach. Econ.
J. 115:28–67 Leighley JE, Nagler J. 2007. Unions, voter turnout, and class bias in the US electorate, 1964–2004. J. Polit.
69:430–41 Levitsky S. 2001. Organization and labor-based party adaptation: the transformation of Argentine Peronism
in comparative perspective. World Polit. 54:27–56 Levitsky S, Mainwaring S. 2006. Organized labor and democracy in Latin America. Comp. Polit. 39(1):21–42 Lindvall J. 2013. Union density and political strikes. World Polit. 65:539–69 Lipset SM. 1983. Radicalism or reformism: the sources of working-class politics. Am. Polit. Sci. Rev. 77:1–18 Marks G. 1989. Unions in Politics. Princeton, NJ: Princeton Univ. Press McAdam D, Tarrow S, Tilly C. 2001. Dynamics of Contention. New York: Cambridge Univ. Press Milkman R. 2006. La Story: Immigrant Workers and the Future of the US Labor Movement. New York: Russell
Sage Found. Moe TM. 1985. Control and feedback in economic regulation: the case of the NLRB. Am. Polit. Sci. Rev.
79:1094–116 Moe TM. 2011. Special Interest: Teachers Unions and America’s Public Schools. Washington, DC: Brookings Inst.
Press Moller S, Alderson AS, Nielsen F. 2009. Changing patterns of income inequality in U.S. counties, 1970–2000.
Am. J. Sociol. 114:1037–101
430 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Mosimann N, Pontusson J. 2016. Enlightenment and solidarity: national union movements, distributive norms and the union effect on support for redistribution. Work. pap., Univ. Geneva
Mosley L, Uno S. 2007. Racing to the bottom or climbing to the top? Economic globalization and collective labor rights. Comp. Polit. Stud. 40:923–48
Murillo MV. 2001. Labor Unions, Partisan Coalitions, and Market Reforms in Latin America. New York: Cambridge Univ. Press
Murphy KJ. 1999. Executive compensation. In Handbook of Labor Economics, ed. O Ashenfelter, D Card, 3:2485–563. Amsterdam: Elsevier
OECD. 2016. OECD employment and labour market statistics. Dataset, OECD iLibrary. http://www.oecd- ilibrary.org/employment/data/oecd-employment-and-labour-market-statistics_lfs-data-en
Piketty T. 2014. Capital in the 21st Century. Cambridge, MA: Harvard Univ. Press Piketty T, Saez E, Stantcheva S. 2014. Optimal taxation of top labor incomes: a tale of three elasticities. Am.
Econ. J. Econ. Policy 6:230–71 Pontusson J. 2013. Unionization, inequality and redistribution. Br. J. Ind. Relat. 51:797–825 Pontusson J, Rueda D. 2010. The politics of inequality: voter mobilization and left parties in advanced industrial
states. Comp. Polit. Stud. 43:675–705 Pontusson J, Rueda D, Way CR. 2002. Comparative political economy of wage distribution: the role of
partisanship and labour market institutions. Br. J. Polit. Sci. 32:281–308 Radcliff B, Davis P. 2000. Labor organization and electoral participation in industrial democracies. Am. J.
Polit. Sci. 44(1):132–41 Robertson GB. 2004. Leading labor: unions, politics, and protest in new democracies. Comp. Polit. 36(3):253–72 Robertson GB. 2007. Strikes and labor organization in hybrid regimes. Am. Polit. Sci. Rev. 101:781–98 Rosenfeld J. 2006a. Desperate measures: strikes and wages in post-accord America. Soc. Forces 85:235–65 Rosenfeld J. 2006b. Widening the gap: the effect of declining unionization on managerial and worker pay,
1983–2000. Res. Soc. Stratif. Mobil. 24:223–38 Rosenfeld J. 2010. Economic determinants of voting in an era of union decline. Soc. Sci. Q. 91:379–95 Rosenfeld J. 2014. What Unions No Longer Do. Cambridge, MA: Harvard Univ. Press Rosenfeld J, Kleykamp M. 2009. Hispanics and organized labor in the United States, 1973 to 2007. Am. Sociol.
Rev. 74:916–37 Rosenfeld J, Kleykamp M. 2012. Organized labor and racial wage inequality in the United States. Am. J. Sociol.
117:1460–502 Rueda D. 2005. Insider–outsider politics in industrialized democracies: the challenge to social democratic
parties. Am. Polit. Sci. Rev. 99:61–74 Saint-Paul G. 2004. Why are European countries diverging in their unemployment experience? J. Econ.
Perspect. 18:49–68 Scheve K, Stasavage D. 2009. Institutions, partisanship, and inequality in the long run. World Polit. 61:215–53 Scholzman KL, Verba S, Brady HE. 2012. The Unheavenly Chorus: Unequal Political Voice and the Broken Promise
of American Democracy. Princeton, NJ: Princeton Univ. Press Scruggs L, Lange P. 2002. Where have all the members gone? J. Polit. 64:126–53 Shin T. 2014. Explaining pay disparities between top executives and nonexecutive employees: a relative bar-
gaining power approach. Soc. Forces 92(4):1339–72 Slaughter MJ. 2007. Globalization and declining unionization in the United States. Ind. Relat. 46:329–46 Sojourner AJ. 2013. Do unions promote members’ electoral office holding? Evidence from correlates of state
legislatures’ occupational shares. Ind. Labor Relat. Rev. 66:467–86 Stepan-Norris J, Zeitlin M. 2002. Left Out: Reds and America’s Industrial Unions. New York: Cambridge Univ.
Press Swenson P. 1991. Bringing capital back in, or social democracy reconsidered: employer power, cross-class
alliances, and centralization of industrial relations in Denmark and Sweden. World Polit. 43:513–44 Teitelbaum E. 2010. Mobilizing restraint: economic reform and the politics of industrial protest in South
Asia. World Polit. 62:676–713 Terriquez V. 2011. Schools for democracy: labor union participation and Latino immigrant parents’ school-
based civic engagement. Am. Sociol. Rev. 76:581–601
www.annualreviews.org • Unions and Inequality 431
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20CH22-Ahlquist ARI 24 March 2017 14:39
Thelen K. 2004. How Institutions Evolve. New York: Cambridge Univ. Press Tope D, Jacobs D. 2009. The politics of union decline: the contingent determinants of union recognition
elections and victories. Am. Sociol. Rev. 74:842–64 Valenzuela S. 1989. Labor movements and transitions to democracy: a framework for analysis. Comp. Polit.
21:445–72 Visser J. 2015. Institutional Characteristics of Trade Unions, Wage Setting, State Intervention and Social
Pacts database version 5.0. Tech. rep., Amsterdam Inst. Advanced Labour Stud., Amsterdam Volscho TW, Kelly NJ. 2012. The rise of the super-rich power resources, taxes, financial markets, and the
dynamics of the top 1 percent, 1949 to 2008. Am. Sociol. Rev. 77:679–99 Voss K, Sherman R. 2000. Breaking the iron law of oligarchy: union revitalization in the American labor
movement. Am. J. Sociol. 106:303–49 Wallerstein M. 1990. Centralized bargaining and wage restraint. Am. J. Polit. Sci. 34:982–1004 Wallerstein M. 1999. Wage setting institutions and pay inequality in advanced industrial societies. Am. J.
Polit. Sci. 43:649–80 Weil D. 1999. Are mandated health and safety committees substitutes for or supplements to labor unions?
Ind. Labor Relat. Rev. 52:339–60 Western B, Rosenfeld J. 2011. Unions, norms, and the rise in us wage inequality. Am. Sociol. Rev. 76:513–37 Wood EJ. 2000. Forging Democracy from Below. New York: Cambridge Univ. Press
432 Ahlquist
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20-TOC ARI 22 March 2017 10:17
Annual Review of Political Science
Volume 20, 2017Contents
Politics, Academics, and Africa Robert H. Bates � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1
Qualitative Methods John Gerring � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �15
Just War Theory: Revisionists Versus Traditionalists Seth Lazar � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �37
International Courts: A Theoretical Assessment Clifford J. Carrubba and Matthew Gabel � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �55
Political Economy of Taxation Edgar Kiser and Steven M. Karceski � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �75
Comparing Political Values in China and the West: What Can Be Learned and Why It Matters Daniel A. Bell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �93
Culture, Politics, and Economic Development Paul Collier � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 111
Progovernment Militias Sabine C. Carey and Neil J. Mitchell � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 127
Voter Identification Laws and Turnout in the United States Benjamin Highton � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 149
Climate Change and International Relations (After Kyoto) Arild Underdal � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 169
Social Movement Theory and the Prospects for Climate Change Activism in the United States Doug McAdam � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 189
Climate Change: US Public Opinion Patrick J. Egan and Megan Mullin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 209
The Political Economy of Regional Integration Christina J. Schneider � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 229
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20-TOC ARI 22 March 2017 10:17
Bureaucracy and Service Delivery Thomas B. Pepinsky, Jan H. Pierskalla, and Audrey Sacks � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 249
Feminist Theory Today Kathy E. Ferguson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 269
When Does Globalization Help the Poor? Nita Rudra and Jennifer Tobin � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 287
Measuring Public Opinion with Surveys Adam J. Berinsky � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 309
Conflict and Cooperation on Nuclear Nonproliferation Alexandre Debs and Nuno P. Monteiro � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 331
From a Deficit of Democracy to a Technocratic Order: The Postcrisis Debate on Europe Ignacio Sánchez-Cuenca � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 351
Understanding the Political Economy of the Eurozone Crisis Jeffry Frieden and Stefanie Walter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 371
The Electoral Consequences of Corruption Catherine E. De Vries and Hector Solaz � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 391
Labor Unions, Political Representation, and Economic Inequality John S. Ahlquist � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 409
Coding the Ideological Direction and Content of Policies Joshua D. Clinton � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 433
Wealth Inequality and Democracy Kenneth Scheve and David Stasavage � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 451
The New New Civil Wars Barbara F. Walter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 469
State Building in the Middle East Lisa Blaydes � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 487
Information, Uncertainty, and War Kristopher W. Ramsay � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 505
Large-Scale Computerized Text Analysis in Political Science: Opportunities and Challenges John Wilkerson and Andreu Casas � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 529
Trading in the Twenty-First Century: Is There a Role for the World Trade Organization? Judith Goldstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 545
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
PL20-TOC ARI 22 March 2017 10:17
Police Are Our Government: Politics, Political Science, and the Policing of Race–Class Subjugated Communities Joe Soss and Vesla Weaver � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 565
Errata
An online log of corrections to Annual Review of Political Science articles may be found at http://www.annualreviews.org/errata/polisci
A nn
u. R
ev . P
ol it.
S ci
. 2 01
7. 20
:4 09
-4 32
. D ow
nl oa
de d
fr om
w w
w .a
nn ua
lr ev
ie w
s. or
g A
cc es
s pr
ov id
ed b
y O
hi o
St at
e U
ni ve
rs ity
L ib
ra ry
o n
10 /2
4/ 22
. F or
p er
so na
l u se
o nl
y.
- Annual Reviews Online
- Search Annual Reviews
- Annual Review of Political Science Online
- Most Downloaded Political Science Reviews
- Most Cited Political Science Reviews
- Annual Review of Political Science Errata
- View Current Editorial Committee
- All Articles in the Annual Review of Political Science, Vol. 20
- Politics, Academics, and Africa
- Qualitative Methods
- Just War Theory: Revisionists Versus Traditionalists
- International Courts: A Theoretical Assessment
- Political Economy of Taxation
- Comparing Political Values in China and the West: What Can BeLearned and Why It Matters
- Culture, Politics, and Economic Development
- Progovernment Militias
- Voter Identification Laws and Turnout in the United States
- Climate Change and International Relations (After Kyoto)
- Social Movement Theory and the Prospects for Climate Change Activism in the United States
- Climate Change: US Public Opinion
- The Political Economy of Regional Integration
- Bureaucracy and Service Delivery
- Feminist Theory Today
- When Does Globalization Help the Poor?
- Measuring Public Opinion with Surveys
- Conflict and Cooperation on Nuclear Nonproliferation
- From a Deficit of Democracy to a Technocratic Order: The Postcrisis Debate on Europe
- Understanding the Political Economy of the Eurozone Crisis
- The Electoral Consequences of Corruption
- Labor Unions, Political Representation, and Economic Inequality
- Coding the Ideological Direction and Content of Policies
- Wealth Inequality and Democracy
- The New New Civil Wars
- State Building in the Middle East
- Information, Uncertainty, and War
- Large-Scale Computerized Text Analysis in Political Science:Opportunities and Challenges
- Trading in the Twenty-First Century: Is There a Role for the WorldTrade Organization?
- Police Are Our Government: Politics, Political Science, and the Policing of Race–Class Subjugated Communities
