The Better Bargain Survey explores workers’ attitudes about their jobs and organized labor; their appetite for greater support, voice, and power in the workplace; and their reactions to political messages and policy reforms

Download Survey Report (PDF)

Executive Summary

The American Compass Better Bargain Survey explores the experiences of American workers and the opportunities for a reinvigorated, modern labor movement to serve them better. In partnership with YouGov, American Compass surveyed a representative sample of 3,000 working-age (18–65) adults on their attitudes about their jobs and organized labor; their appetite for greater support, voice, and power in the workplace; and their reactions to political messages and policy reforms.

  • Excessive engagement in politics is the number one obstacle to a robust American labor movement.
    • By a three-to-one margin, potential union members (part- or full-time employees who work 30 or more hours per week at private, for-profit companies in non-supervisory roles) say they would prefer a worker organization that focuses only on workplace issues to one that is also engaged in national political issues.
    • Asked how important various activities are for a worker organization to pursue, potential union members assigned half their weight to collective bargaining and providing training and benefits, versus 3% to politics.
    • Only 35% of potential union members not already in a union say they would vote for one—roughly equal to the 34% who are undecided and 31% who say they would vote no. The top reason for voting no is union involvement in politics; by contrast, fear of retaliation is the reason least cited for voting no.
    • Across nearly 20 political issues highlighted by the AFL-CIO and SEIU, not a single one was chosen by a majority of potential union members as something they would want to see a worker organization focus on; most issues were chosen by one-fifth of workers or fewer.
  • The second obstacle to a robust American labor movement is the unfamiliarity of most workers and their apathy toward its existing form.
    • Americans generally have favorable views of labor unions, but it is middle- and upper-class Democrats whose views are most favorable; lower- and working-class views (across parties) are less favorable, and those workers are much more likely to say they just “don’t care.”
    • Most lower- and working-class Americans also say they either want politicians to ignore labor unions or don’t care what politicians say; when it comes to specific organizing efforts though, the vast majority of workers say that politicians should support the effort, including majorities of lower- and working-class Republicans.
    • When asked why their own workplaces are not unionized, two-thirds of potential union members say either that they are not sure or that the employees have never thought about it; only 2% cite fear of employer retaliation.
    • When asked what forms of cooperation between worker organizations and management are desirable or necessary, the most common answer from potential union members is “don’t know.”
  • Workers’ attitudes toward their own jobs and experiences in the workplace define the contours for how effective reform should look: cooperative, beneficial to labor-management relations, and focused on concrete economic benefits.
    • Fewer than one-in-three workers has a “secure” job, defined as annual income of at least $40,000 with predictable earnings, steady hours, and health benefits; among workers without college degrees, that figure falls to one-in-five.
    • Most workers have never negotiated with their employer to obtain a significant change to their compensation, benefits, or any other term or condition of employment.
    • The key driver of satisfaction at work is not objective measures of job quality but rather the character of labor-management relations; workers who report an “Excellent” relationship with management are highly likely to recommend their job to a friend and look forward to going to work, while the opposite is true for those who report a “Poor” or “Fair” relationship.
    • Likewise, most workers who report “Excellent” labor-management relations are content dealing directly with management themselves on workplace issues and negotiating their own terms of employment, while this falls to 50% for those with “Good” relationships and to just one-third for those with “Poor” or “Fair” relationships.
    • Workers prefer a worker organization run jointly by management and employees to one run by employees alone (63% to 37%) and want to see their representatives chosen from among employees rather than from an outside organization (77% to 23%); their priorities for an organization’s activities are (1) collective bargaining, (2) benefits and training, and (3) workplace collaboration.
  • Policymakers and labor leaders seeking to address the concerns and interests of working-class Americans broadly and potential union members in particular should orient their messages and reforms toward a new conception of the labor movement.
    • The messages favored by Big Labor, Big Business, and progressive academics predictably perform well with some partisans and poorly with others, but an emphasis on “new unions” that “provide services to workers and represent them in the workplace without dragging down businesses or getting involved in politics” is popular across the board and with Independents especially.
    • Involvement in politics and social activism turns workers off to the efforts of organizers and the project of collective worker power generally, while the provision of concrete benefits and opportunities for collaboration in the workplace are surefire winners.
    • One policy proposal that earns especially broad support is a repurposing of the labor organization’s role: prohibiting political spending but allowing funding from employers and governments to provide training programs and benefits; Americans across classes and political parties express favorable views by wide margins.


Part I: How Americans Feel About Unions

Most Americans have favorable attitudes toward unions in the abstract, but they are not especially interested in hearing politicians talk about them and not inclined to vote for one themselves. Their main objection is union involvement in politics.

I.A General Attitudes

Most Americans have a favorable view of labor unions, though attitudes diverge predictably along partisan lines. Democrats are +64% favorable while Independents are +20% and Republicans are -12%.

While organized labor might seem primarily a concern of the lower and working classes, in America it is the middle and upper classes who have stronger and more strongly divergent attitudes. Middle- and upper-class Democrats have much more strongly positive views than lower- and working-class Democrats. Among Republicans, the middle and upper classes have strongly negative views while lower- and working-class views are closer to neutral.

Across parties, lower- and working-class adults are also more likely to be apathetic towards unions, saying they just “don’t care.” Among Democrats, 25% of lower-class adults “don’t care” compared to 2% of upper-class adults; among Republicans, those figures are 22% and 8%, respectively.

I.B Political Rhetoric

This pattern of broadly favorable attitudes combined with lower- and working-class apathy appears in political preferences as well. Americans are nearly four times more likely to prefer that politicians speak favorably rather than unfavorably about labor unions. But majorities of lower- and working-class adults either don’t care what politicians say about unions or would prefer that they not speak about unions at all.

Only among Democrats does a majority of any class want politicians to speak favorably about labor unions, with the upper and middle classes expressing the strongest preference. Republicans across classes are least likely to want to hear politicians speaking favorably, but lower- and working-class Republicans prefer speaking favorably to speaking unfavorably.

I.C Organizing at Work

Americans’ generally favorable views toward labor unions do not translate into enthusiasm for joining one.

In this analysis, we define “potential union workers” as part- or full-time employees who work 30 or more hours per week at private, for-profit companies in non-supervisory roles. These workers are closest to the archetypal private-sector employee and represent the subset of the labor force most likely to be served by a conventional labor union.

Among potential union members, 37% say they would vote to join a union, 29% say they would be opposed, and 33% say they are undecided.

But of course, some already are union members. Among those not already in a union, which offers a better view of prospects for new organizing activity under the current system, the 35% of potential union members saying they would vote to join is indistinguishable from the 34% undecided and 32% opposed.

Asked why their workplaces are not unionized, more than two-thirds of potential union members say that they either are not sure or have never thought about it, whereas fewer than one-in-six cite concerns related to their employer. While the conventional narrative from labor organizers holds that fear of employer retaliation deters efforts at organizing, only 2% of potential union members cited this factor.

I.D Union Political Involvement

While American workers may be unsure about organizing in their own workplaces, there is one labor issue about which they are absolutely clear: they don’t like union involvement in politics.

Among those who said they would vote against a union, the top reason cited was union political activity, followed by member dues. These workers anticipate that unions will focus on politics rather than delivering concrete benefits in their workplaces, and don’t want to pay the cost. Once again, retaliation finished a distant last in explanations for hesitancy about organizing.

But surely there must be some political issues that most workers would want to see their organization engaged in? Not so.

The survey presented a list of 19 different issues that the AFL-CIO and SEIU highlighted as priorities on their websites as of July 2021, and asked potential union members to select whichever ones they would like to see their own hypothetical worker organization speak out on.

Workers were most likely to pick bread-and-butter labor issues like workplace health and safety, minimum wages, and Social Security and retirement, but not even those were chosen by a majority of workers. Most issues were selected by one-fifth or fewer.

Unsurprisingly then, workers say what they really want is to have unions avoid politics altogether.

Asked whether they would prefer to be a member of a worker organization that devoted some of its resources to national political issues or one that focused on workplace issues only, the American labor force choose the latter by a two-to-one margin.

Among potential union members, that margin rises to three-to-one.

A distaste for politics in the workplace also goes beyond unions. As more corporations have taken public stances on behalf of progressive political causes, party politics risks becoming a workplace issue that workers would rather avoid.

59% of the American labor force and 63% of potential union members say that they would prefer for their employers to stay out of social justice issues.


Part II: Satisfaction and Voice in the Workplace

Relations with management are more important for workers’ job satisfaction than objective measures of job quality, but workers are conflicted about how to engage with managers in negotiations and the resolution of workplace issues. Most workers prefer organizations with different priorities, structures, and relationships with management than what existing labor unions offer.

II.A Job Quality

Most American workers have insecure jobs.

In this analysis, we define a “secure job” as any work that pays a minimum of $40,000 per year for a minimum of 30 hours per week, with stable earnings, predictable hours, and health benefits. All jobs that don’t meet these criteria are “insecure jobs.”

Only 30% of American workers have secure jobs, and they are concentrated among workers with college degrees. 45% of workers with at least a 4-year college degree have a good job, compared to just one-in-five workers without a degree.

Even among the archetypal private-sector employees defined here as “potential union members,” only 43% have secure jobs, and this rises to a majority only for those with college degrees.

American workers do report a better situation when it comes to the more subjective question of relations with management.

Asked how they rate relations between employees and management at their workplace, 20% of workers say “Excellent” and an additional 48% say “Good.” Among potential union members, those figures are 21% and 45%, respectively. Fewer than one-in-ten report “Poor” relations with management, regardless of class.

While the typical American worker gets along well with management, he isn’t likely to be negotiating on his own behalf. Only one-third of workers, and one-third of potential union members, report successfully requesting and receiving a significant change in their pay, benefits, schedule, or some other term or condition of employment in the last five years.

Outside of the upper class, a majority of workers have never negotiated successfully.

II.B Net Promoter Score

How satisfied are Americans with their jobs?

The survey asked workers how likely they would be to recommend their job to a friend, on a scale from zero (“not at all likely”) to ten (“extremely likely.”) Those who respond with a nine or ten are considered “promoters,” who are highly satisfied with their jobs, while responses between zero and six indicate “detractors,” who are likely unsatisfied. The difference between the shares of “promoters” and “detractors” creates a Net Promoter Score (NPS), commonly used by organizations in assessing their own employees’ satisfaction.

Scores varied little between American workers generally (-9) and potential union members specifically (-11), but there was significant variation by class.

Lower- and working-class workers both gave their jobs scores of -15, whereas upper-class workers gave theirs a score of +12.

These scores are most useful in making comparisons across potential drivers of job satisfaction. As would be expected, scores vary by objective measures of job quality: “Secure jobs” earn a score of +4 and “insecure jobs” a score of -15. But compare that result to the divergence by self-reported quality of labor-management relations.

Workers who report having “Poor” or “Fair” relations with management give their jobs scores of -94 and -59, respectively, compared with a score of +3 for jobs with “Good” relations and +55 for those with “Excellent” relations.

Each step down in quality of relations has an effect more than twice as large as the effect of moving from “secure” to “insecure” jobs.

II.C How to Relate

When handling a workplace issue, roughly half of workers and potential union members say that they would prefer to deal with management directly themselves, rather than deal together with their fellow workers to address it. Roughly a fifth are unsure, and just less than a third prefer collective action.

Preferences vary modestly by class, but somewhat counterintuitively: it is only upper-class workers who prefer collective to individual action, 43% to 36%.

Despite a lack of experience negotiating successfully on their own, roughly half of workers and potential union members say that they would prefer to negotiate with management directly, while one-third would prefer dealing together with fellow workers.

Again, views vary by class, with upper-class workers most likely to say they would prefer joint versus individual negotiations—by 46% to 33%.

But workers’ attitudes about dealing with management depend on how well they get along in the first place. As with job satisfaction, the preference for individual negotiation and arbitration is correlated with better workplace relations.

II.D Ideal Worker Organization

American labor law and labor organizations give workers few options. What workers say they want is not what they get.

Our survey asked respondents to allocate 20 points across a list of seven different activities for a hypothetical worker organization, based on how important they considered each. Potential union members allocated most of their points to bread-and-butter tasks and almost none to politics. Collective bargaining, benefits, and training received nearly half of the total; politics and social activism together received just 8%.

Workers also want very different structures in a labor organization than what today’s unions provide, with much greater employer collaboration and local autonomy.

By a three-to-one margin, potential union members say they would prefer to be represented by fellow employees rather than by someone from an outside organization.

Likewise, potential union members prefer local funding sources for a worker organization. Asked to choose as many sources of funding as they liked, 63% of potential union workers support member funding and 43% say that employers should contribute to the costs. Less than a fifth of potential union members say that a nonprofit or government agency should supply funding.

On questions of cooperation versus conflict with management, workers’ views are more mixed and less well formed.

Asked whether management cooperation is necessary for a successful worker organization, potential union members are evenly split between 27% who say that cooperation is necessary and 27% who say that it is unnecessary.

A higher share, though, say they don’t know.

Workers are likewise unsure about trade-offs between worker power and labor-management cooperation.

When presented the choice between a worker organization that lacks power but has management cooperation or one with more power that management opposes, nearly half of potential union members say that they don’t know or don’t prefer either. 24% choose power without support and 15% choose cooperation without power.

This result differs from the finding when the question was first polled more than two decades ago, in the What Workers Want study. Then, nearly two-thirds of workers preferred an organization with management cooperation but no power to make decisions. Attitudes have not shifted strongly to a preference for power at the expense of management support, but the trend suggests that employers have been losing the trust of workers to act cooperatively.

Hypothetical questions of power and cooperation are notably abstract, so the survey also presented workers with a concrete choice about organization structure. When forced to choose between the options presented, worker attitudes remain strongly inclined toward the cooperative model.

Nearly two-thirds of potential union members say they would prefer a worker organization run jointly by employees and management to one run by employees alone. Similar shares of those unsure about management cooperation and worker power in theory also prefer a jointly run organization in practice.


Part III: Political Messaging and Policy Reform

Most Americans support reforms to get unions out of politics and refocused on activities like providing training and benefits, and most would like to see politicians support workers’ organizing efforts. Political rhetoric stressing modern reform and a constructive role for unions garners broad appeal across the political spectrum.

III.A Benefits

Workers have widely varying views on their preferred benefits providers. For instance, pluralities of the lower, working, and middle classes prefer employer-sponsored health care. But that model fails to win majority support from any class. Compared to upper- and middle-class workers, lower- and working-class respondents are more likely to prefer shopping for coverage themselves or receiving it through a worker organization.

A similar pattern emerges for unemployment insurance. Workers across all classes prefer the status quo of unemployment benefits administered by state governments, but most lower- and working-class workers would prefer a different model, whether through a worker organization or former employer.

III.B Better Bargains

Despite conflicting views on the ideal benefits provider, shifting union activity away from politics and toward a Ghent-style system of benefits is an overwhelmingly popular policy proposal with the American public.

The survey presented the following policy reform: “Unions would no longer be allowed to spend money on political campaigns or causes, but they could receive funding from employers and the government to set up training programs and provide health and unemployment insurance benefits to workers.”

Majorities of American adults across all classes support the proposal, which had overall net favorability of +36%.

The proposal also garners bipartisan approval, winning the support of Republicans, Independents, and Democrats—in all cases by a three-to-one margin.

Few policy proposals are this universally popular.

Conversely, Americans are less enthusiastic about a proposal in which workers lose their right to strike but gain the right to elect worker representatives to corporate boards of directors.

Responses are still more favorable than unfavorable, but only by +7%.

The proposal does not win majority support among any class, and many people are simply unsure what to think of the idea. Roughly a third of lower- and working-class adults say they don’t know whether they view the policy proposal favorably.

As with the proposal concerning politics and benefits, this proposal generates quite similar responses across political parties, but again with much lower levels of support. Net favorability among Democrats and Independents is -1% and -6%, respectively.

III.C Other Political Attitudes

Americans have mixed feelings about politicians speaking about labor unions (see Part I.B), but when it comes to specific efforts at organizing, they want to see support.

Nearly two-thirds of Americans think that politicians should publicly support workers’ organizing efforts compared to only one-tenth who think that politicians should oppose them.

Attitudes here are partisan: 85% of Democrats would like politicians to support organizing efforts, compared to 45% of Republicans. But among Republicans, that figure is still higher than the 18% who want politicians to oppose organizing efforts.

Attitudes vary little by class among Independents and Democrats, but markedly among Republicans. 56% of lower-class Republicans and 50% of working-class Republicans want politicians to support workers attempting to organize; 55% of upper-class Republicans prefer that politicians ignore it.

Most Americans feel similarly about public- and private-sector unions, but the quarter who do see a distinction favor private-sector unions by two-to-one over public-sector unions.

III.D Political Messaging

What, specifically, do Americans want to hear their politicians say about labor policy?

Our survey presented four different messages, each representative of a different policy approach, and asked whether each made respondents more or less likely to support a political candidate who said it.

  • Big Labor: “A union card is the single best tool working people have in the fight to win a fair share of the value they create every day.” (This statement comes from a speech delivered by former AFL-CIO president Richard Trumka on the Protecting the Right to Organize (PRO) Act in March 2021.)
  • Big Business: “Less government intervention in the workplace empowers workers and job creators and allows them to prosper. When employees are free to make decisions with their employers, American workers win and the economy grows.” (This statement comes from the website of the anti-labor Institute for the American Worker.)
  • Progressive Values: “Unions are important, but they need reform. This must start with inclusion to ensure that all workers can build power and to address systemic racial and gender oppression.” (This statement comes from the executive summary of Harvard Law School’s Clean Slate for Worker Power report.)
  • Worker Focus: “Workers need new unions for the 21st century that will provide services to them and represent them in the workplace without dragging down businesses or getting involved in politics.” (This statement was drafted by American Compass.)

Each statement was scored based on the percentage of respondents saying it would make them more likely to support a politician minus the percentage saying it would make them less likely. Respondents could also answer “no difference” or “unsure.”

Partisan responses to the “Big Labor” and “Big Business” statements mirror each other, with Independents slightly preferring the “Big Business” approach. “Big Labor” scores +36 among Democrats and -25 among Republicans. “Big Business” scores +46 among Republicans and -20 among Democrats; it is the only statement with a negative score among Democrats. Independents give positive scores to both statements: “Big Business” scored +13 and “Big Labor” +3.

The “Progressive Values” statement performs worst. It scores +36 among Democrats, comparable to “Big Labor.” But it is the least popular statement among Republicans (-26) and the only statement with a negative score among Independents (-5).

The “Worker Focus” statement looks quite different: it does not generate the intense enthusiasm of the other statements with particular partisan groups but its popularity is uniquely widespread. It garners near-equal positive scores from Republicans (+13) and Democrats (+12) and scores better than any other statement among Independents (+22).

American workers are ready for a better bargain on labor, and it will have to look quite different from what either Big Business or Big Labor prefers.



The American Compass Better Bargain Survey was conducted by YouGov between July 16 and 26, 2021, with a representative sample of 3,000 adults aged 18–65 living in the United States. YouGov interviewed 3,355 US adults between ages 18–65 who were then matched down to a sample of 3,000 to produce the final dataset. The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2019 American Community Survey (ACS) one-year sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file).

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, years of education, and region. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The weights were then post-stratified on 2016 and 2020 presidential vote choice, and a four-way stratification of gender, age (3 categories), race (4 categories), and education (4 categories), to produce the final weight.

Respondents were instructed:

As you may be aware, American workers have the right to form labor unions and bargain collectively (as a group) with their employers, and most can go on strike if they are dissatisfied with their working conditions. Some other countries have unions much like the American ones, while others have different types of worker organizations that bargain collectively and perform other duties—such as providing services like health insurance and unemployment benefits, consulting with management on workplace issues, and electing representatives to corporate boards of directors.

This survey explores Americans’ experiences in the workplace, how they feel about today’s labor unions, and what options they would ideally want workers to have.

One technical term you will see is “terms and conditions of employment,” which refers to the different elements of the relationship between workers and employers that they might bargain over: wages, benefits, and hours of work, of course, but also how the workplace is run, how decisions get made, how disciplinary issues are handled, and so forth.

To distinguish between workers at different levels within an organization, the survey uses these terms: 

  • Non-supervisory workers: those who do not oversee the work of other people and do not have the power to hire and fire employees.
  • Managers: those who do supervise other people and often have the power to hire and fire.
  • Owners and Executives: those who either own or manage large parts of an organization and make key business decisions.

“Class” is defined by education and income:

  • “Lower” (N=702): less than a four-year degree and household income below $30K; or did not report household income and do not have a high-school diploma.
  • “Working” (N=950): less than a four-year degree and household income $30K-$80K; or did not report household income and have either a high-school diploma or some college but no degree.
  • “Middle” (N=980): four-year degree or more and household income $30K-$80K; or household income $80K-$150K; or did not report household income and have a 2-year or 4-year college degree.
  • “Upper” (N=258): household income above $150K; or do not report household income and have a post-graduate degree.

Respondents with a four-year college degree or more but household income below $30K are excluded from analyses using the “Class” variable.

Recommended Reading
New Survey Finds Modern American Workers Want Unions To Stop Playing Politics

American Compass executive director Oren Cass joins The Federalist Radio Hour to discuss the Better Bargain Survey results, what workers want from organized labor, and working-class perspectives on politics.

New Survey Finds Union Political Activism the #1 Obstacle to U.S. Labor Movement

PRESS RELEASE—Labor policy reforms should be focused on strengthening labor-management cooperation and delivering concrete economic benefits

Talkin’ (Policy) Shop: Worker Voice

On this episode, Oren and Chris discuss the issue of worker voice and how to reform labor law to create new avenues for labor input from the shop floor to the boardroom.