A couple of days ago another on-demand startup closed its doors. You’ve probably never heard of Washio. It was an on-demand laundry service (yes, really), and only plied its trade in a couple of American cities.

I bring this up because the demise of Washio coincided with the Twitter reemergence of a great Kevin Roose article from back in 2014. Writing at the beginning of the on-demand boom, Roose marvelled at the torrent of service-based start-ups. Many featuring non-existent business models, betting big that these markets were only small because of price sensitivity and “friction”.

This business model is great for consumers. As a result of start-ups’ willingness to lose money for months or years at a time, I get cheap, fast services that come with an effective subsidy that can add up to thousands of dollars a year. But they’re problematic for the businesses themselves. Unlike Amazon or Google (which have profitable core operations that subsidize the money-losing services elsewhere in their business), or Uber (which uses the profits from its high-margin Uber Black and Uber SUV lines to subsidize its low-margin UberX service), many of today’s start-ups have no profitable parent company pouring in money. They’re simply taking millions of dollars in venture capital with the hope of keeping prices low, pushing rivals out of the market, and eventually finding a way to turn a profit.

But there’s another great point buried in the Roose article. Where this money is coming from. It’s just so easy to talk about “venture backed” start-ups without considering where the money in these funds come from. A big source is pension pools. The pensions of many people who have little access to the services they are subsidising – they don’t live in the few big cities these start-ups called home.

while some of the money used to fund money-losing start-ups comes from rich Silicon Valley investors, some large amount of it comes from public pensions, college endowments, and other, more modest sources. Lyft backer Andreessen Horowitz, for example, has gotten investments from the Imperial County, California, Employee Retirement System and the University of Michigan; the Tennessee Consolidated Retirement System invests money with SpoonRocket backer General Catalyst. If you asked them, I’m sure that firefighters in Memphis and public schoolteachers in El Centro would have no idea that their retirement funds are being used to lower the price of my delivery lunches and rides across town. But that’s exactly what’s happening. And when these venture-backed price wars happen in dozens of high-end service sectors all at once, you have a strange cultural phenomenon in which Main Street dollars are being used to finance the lifestyles of cosmopolitan yuppies.

(Emphasis added)

It’s probably wrong to think that rural firefighters are losing money because of the bets placed on on-demand start-ups that never worked out unit economics. Their money is only fraction of a huge pool. The money allocated to VCs would be only part of a diverse portfolio. And these VC funds themselves factor in a certain amount of bad bets.

But as the Internet hollows out the middle, as globalisation rebalances industry away from places like Memphis, it’s worthwhile thinking about where the capital is coming from. And whether they are the ones benefitting. This is a rather weird transfer.

You really should read the entire Roose article.