In this photo taken Tuesday, Dec. 16, 2014, a man leaves the headquarters of Uber in San Francisco. (Eric Risberg/AP)
Opinion writer

Flexible hours. Being your own boss. The glories and self-bootstrapping pride of entrepreneurship.

These are among the virtues of “sharing-economy” gigs, as touted in a recent Uber-commissioned survey of its drivers. Other companies offering peer-to-peer platforms, such as Airbnb, TaskRabbit and Homejoy, have made similar pitches: They’re giving workers — particularly those who are unable to land traditional jobs or unfulfilled by 9-to-5 Organization Man duties — the freedom to take their breadwinning fates into their own hands.

It’s true that, in many ways, sharing-economy jobs can offer more autonomy than traditional employer-employee relationships. But there’s a dark side to these work arrangements that gets considerably less press: the shifting of risk off corporate balance sheets and onto the shoulders of individual Americans, who may not even realize what kinds of liabilities they’re taking on.

The risks involve everything from income instability (the worker, rather than the firm, has to absorb the brunt of demand shocks or price cuts); to irreversible capital investments (Uber and Lyft have infamously pushed drivers to buy new cars by promising big returns that never materialized); to unforeseen criminal liabilities (what happens if an Airbnb guest turns your home into a brothel?); to fewer protections in the event of catastrophe (no access to programs such as workers’ comp). Sure, sharing-economy “entrepreneurs” can get a lot of upside, but there are a lot of hidden downsides, too.

Celebration of these riskier arrangements can seem especially strange when you consider that society’s ability to better manage risk, and spread it over larger pools of people, is considered by many historians to be one of the great advances of 20th-century finance. This achievement arose partly because economists developed a much more sophisticated understanding of insurance market design. But it also stemmed from social necessity. The safety nets humans relied on for centuries — their extended families — became less reliable in the age of industrialization and urbanization. As kinship networks frayed, European governments developed robust welfare states. Here in the United States, for reasons driven partly by ideology and partly by historical accident, these new safety nets were largely administered through employers (for example, health insurance). Some historians call this “welfare capitalism.”

Then, beginning around the 1970s, this form of corporate-based risk-sharing began to unravel. Exactly why is debatable; globalization, the decline of unions, regulatory changes, new technology and financial markets all likely played a role. The result, though, is that programs such as defined-benefit pensions began to disappear. Just-in-time scheduling, outsourcing and other arms-length relationships between firms and workers blossomed. In some ways, these developments were very good for economic growth, but they also introduced much more instability into the lives of middle-class workers.

In this context, sharing-economy jobs look a little less revolutionary and more like a logical extension of longer-term trends.

It’s easy, but probably unproductive, to feel nostalgic about the good ol’ days of welfare capitalism. The disintegration of the corporation-centered safety net looks likely to continue. The challenge is to develop policies that mitigate some of the greater risks and sources of instability facing workers, whether as Lyft drivers or temps, especially since we know from behavioral economics that individuals tend to be really bad at managing risk on their own.

The private sector has made half-hearted attempts to offer new insurance products to help sharing-economy “entrepreneurs” deal with precarious work arrangements. Peers.org, for example, recently started selling ridesharing drivers access to a spare car if theirs is in the shop. But as long as such products remain optional, adverse selection and “death-spiraling” seem unavoidable. (Only people with the shoddiest cars may buy the insurance, causing the price to rise, pricing more people out until nobody is left.) And as we’ve learned from Obamacare, individual insurance mandates — basically the only way to guarantee insurance markets don’t death spiral — are not exactly politically popular.

That said, Obamacare is actually one of the few major policy developments of recent decades to mitigate the risk-shifting trends I’ve described. Perhaps it’s not surprising, then, that Uber CEO Travis Kalanick supports it, given that a functional individual health insurance market makes becoming a sharing-economy “entrepreneur” much more sustainable.

I’ve given Uber a hard time about many of its policies, but on this we agree: As long as safety-net programs are being decoupled from the employment contract — and it’s not obvious to me that that was the best place for them to begin with — America probably needs a more robust government safety net to help workers deal with the fallout.