President Donald Trump speaks at Shell's Pennsylvania Petrochemicals Complex on August 13.Susan Walsh / AP

When President Donald Trump arrived in western Pennsylvania this month to give a speech about energy policy at a Royal Dutch Shell plant, he had a ready-made audience composed of workers who, it turns out, were paid to be there. The company suggested that the event was simply a “training day” featuring an unusually prominent guest speaker, and offered that workers could take a day of paid time off instead of attending, which would mean they would lose overtime pay. That alternative may have been realistic for some workers, but others must have felt that the only option was to appear as a living backdrop for the president.

This example is unusually high-profile, but politics and work often come together to workers’ detriment. This dynamic can take different forms, but all variations on the theme raise similar problems: Employers have a largely unconstrained ability to try to influence their workers’ political choices.

The political scientist Alexander Hertel-Fernandez has found that many workplaces are saturated with political messages. Sometimes employers and their lobbyists hope to benefit from workers’ legitimacy on issues that affect them by leveraging their voices in lobbying campaigns. A recent “astroturfing” campaign to oppose a Washington, D.C., initiative to eliminate the tipped minimum wage is one example. Another, older example involves a 1959 law that imposed new constraints on unions. Senator Paul Douglas read into the Congressional Record a constituent letter, which the writer had sent to cancel out an earlier message of support for the law—that one sent at the direction of her boss. “I wonder how many other stenographer-secretaries have spent days and days writing and wiring Senators and Congressmen urging support … in the names of various firm executives, their friends, relatives, and employees,” the letter mused.

Employers also seek to influence their employees’ support for candidates or issues within the workplace, sometimes through coercion or manipulation. Employers have outright fired workers for their political bumper stickers. Others have tried to change their workers’ minds; for example, the CEO of the time-share company Westgate Resorts sent his 7,000 employees an email predicting layoffs and benefit cuts if President Barack Obama was reelected in 2012. Was he serious? Apparently not, but it is easy to imagine workers thinking otherwise when casting their ballots.

Federal law has very little to say about employers who exert these kinds of pressures at work; for example, no federal law prohibits private-sector employment discrimination based on political viewpoints. Some state and local laws pick up part of the slack, but they tend to be limited to the most coercive employment practices, such as threatening to fire employees because they expressed support for a particular candidate.

One reason legislatures do not do more is the First Amendment—specifically, the Supreme Court’s robust protection for corporate political advocacy. In 2010, five conservative justices agreed in Citizens United v. FEC that “if the First Amendment has any force, it prohibits Congress from fining or jailing citizens, or associations of citizens, for simply engaging in political speech.” That lofty language translates to a First Amendment right for corporations to spend unlimited general treasury funds on political advocacy. It also means that a law targeting employers’ noncoercive attempts to influence their employees’ politics would face an uphill climb.

Although Citizens United applies to unions as well as corporations, unions may not require represented workers to contribute to their political advocacy. No equivalent funding limitation applies to corporations; for example, shareholders cannot opt out of a corporation’s political spending.

A combination of lack of political will and the First Amendment means that we are unlikely to see more robust legislative responses to reduce employers’ political influence. But some unionized employees are better positioned than others to resist pressure.

First, union contracts tend to bring better job security; for example, most contain “just cause” provisions, which mean employees can’t be fired without a good reason. To state the obvious, “I didn’t like their political bumper sticker” and “They didn’t promise to vote for my candidate” are not good reasons.

Second, unions can talk politics with workers—though they cannot visit employment consequences on those who decline to engage—potentially offering a point of view that is different than that of the employer. Unions often also encourage their members to vote and get involved in politics in other ways; one study found that “union members, particularly those without any college education, are significantly more politically knowledgeable than their non-union counterparts.”

Unions cannot prevent employer politicking altogether, of course. The Shell employees were mostly unionized, and yet they still faced a choice between attending Trump’s speech and losing part of their compensation. But things may have been worse if the employees were nonunion: In 2012, miners were ordered by their managers to give up a day’s pay to attend a Mitt Romney campaign event.

One way, then, to decrease the weight of employers’ political pressure on employees is to increase unionization. Government could also replicate one of the benefits of union representation by enacting just-cause legislation. But just as the Supreme Court stands in the way of some limits on employers’ political influence over their workforce, it has also weakened unions.

As mentioned above, unions have been unable for decades now to require represented employees to pay the portion of union dues that the union uses for political advocacy. And last year, the Supreme Court held in Janus v. AFSCME that public-sector workers cannot be required to pay even their share of union collective-bargaining costs, because workers might object to unions’ bargaining positions.

Janus applies only to public-sector unions, but about half of American union members are public employees. In addition, about half of states have right-to-work laws, which apply in the private sector. So if you are a public-sector worker anywhere, or a private-sector worker in a right-to-work state, you can’t be required to pay for union representation. A predictable result of both Janus and right-to-work laws is to dilute unions’ political strength: Both in effect require unions to stretch dues to cover representation of nonmembers before they can spend on politics.

The 2020 election cycle is already under way; many employees will be subjected to political pressure at work in the coming 15 months. Effective responses would involve stricter limits on employers’ politicking at work—limits that recognize that employers’ political pressure tends to be inherently coercive—more job protections for nonunion workers, and stronger unions. There’s no reason to believe that the first two responses will materialize, at least not at the federal level. That leaves stronger unions, a possibility that rests in workers’ own hands.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.