26 Apr 2022 Unionization In the Time of Automation

Unions have recently won organizational votes at New York City-area branches of Starbucks and Amazon, companies they have been trying to crack into for years.

Scott Shepard

There is good reason to think that workers enticed into voting for unions in the 2020s will come to regret their decision. Modern unions are generally bad for economies, bad for polities, bad for companies, bad for employees and bad for civic life generally. They seem particularly ill-fitted for the current moment, as both the Starbucks and Amazon examples illustrate.

In a fully free society, people would be free to join any organizations they wanted. Likewise, employers would be bound only by the contracts that they signed, so that employers who hired workers on an “at-will” basis, without a contract, could fire those employees if they joined an employees’ group that started to make demands the owner didn’t like. What we in the United States have is not that fully free society, but rather one in which employer rights have been statutorily constrained for the benefit of unions, most especially in the Wagner Act of 1935 and subsequent legislation. Before that act, unions were broadly either helpful or harmless. After the elevation of the unions to superior status, they became a potential menace. They lined the pockets of union executives while slowing the pace of work, decreasing productivity, charging employees and thwarting the efforts of entrepreneurs of mind (e.g., start-up companies) or of muscle and hustle (e.g., workers who wished to succeed up to the level of their ability, rather than to live down to the constraints of union-demanded work rules).

By the 1970s, though, these unions were harming American productivity and – along with Dick Nixon and Jimmy Carter and Art Burns and a whole cast of catalysts – the American economy. The case was even worse, and more stark, in Britain, where the post-war socialist governments had given unions such power that they were able – and happy – to strangle the life out of the British economy for the benefit of unproductive and money-losing industries like state-owned coal mines, until the country was saved and the unions brought to heel by the Thatcher Renaissance.

And here we are again. It is heartening to see that so far workers in other parts of the country have been resisting unionization at Amazon and Starbucks and in general. Unionization worked out poorly – including, in the long term, for all those employees who found themselves unemployed because their unions had made the companies they worked for unprofitable – in the 1970s. They will be worse for workers now.

Automation growth, which famously accelerated in response to the lockdowns, remains high. Ever more jobs that were once filled by young or low-skilled workers have been automated, and many more can be in coming years. The higher unions drive wages in labor-intensive, low-margin industries, the more tasks become cheaper to automate. Union officials want to unionize regardless, because more members mean more paychecks from which to extract union dues. For employees, though, the calculation is much more complicated. For them, unions mean the possibility of a wage boost at some point, but the certainty that whether wages rise or not, the union will be taking dues out of those paychecks. So the possibility of some gain is paired with the guarantee of certain loss. And then there’s the consideration that the gain can’t be very large, because if it were the company would either be crushed by labor costs, closing stores or going under entirely; or would automate workers out of their jobs.

This dynamic plays out slightly differently at Amazon than at Starbucks. At Amazon the vital threat to workers is automation. Already the company is building “dark” warehouses (sorry, “fulfillment centers”) for its Whole Foods subsidiary – dark because they’re worked entirely by robotic equipment that doesn’t need lighting. Those can easily be rolled out companywide. Once labor costs get higher than automation costs, they will be. That’s the ceiling for warehouse-worker wages, full stop. Given union propensity to make employees less productive, the primary effect of a union at Amazon warehouses will be to lower the value and raise the cost of workers, speeding up the automation timeline.

For Starbucks workers, meanwhile, the initial calculation is one of longevity. Fast-food workers don’t generally stay at those jobs for very long. They’ll have to pay union dues out of their paychecks for the limited time they work at Starbucks, but are unlikely to accrue much overall benefit from their membership for the time that they are there. Rather, they will subsidize the pay and perks of union officials who won’t much care about them, since they’ll be gone soon. And as any benefits unions achieve for workers will also come with the productivity drag of unions, the higher-costs/lower-productivity combination will mean marginal hours will be cut and stores will be closed, costing future potential workers any opportunity at a job at all.

On top of all this, there is the fact of what unions do, and are. Employees are courted by union activists with stories of a golden future in which employers will get off their backs while spiking their pay. As we’ve seen, though, that pay spike, if it comes, doesn’t – can’t – last for long. Equally flawed is the notion that having a union means additional freedom. It doesn’t. It means another set of people who are establishing rules you have to follow and whose favor you have to court (i.e., people on whose good side you must stay) in order to get any support or redress. The goal of greater freedom is not advanced by voluntarily donning the harness of an additional set of masters. Go watch On the Waterfront. It’s nearly 70 years old, but human nature is – contra lunatic woke doctrine – eternal. And so the nature and effects of statutorily privileged unions have not changed.