Steve Randy Waldman asked that question this year, and he raises some great points about what is a generally accepted microeconomics tenet. As a reminder, a market is said to clear where the demand and supply curves meet.

Here at interfluidity, we are not in the business of useless economics, so we will adopt a very conventional utilitarianism, which assumes that people derive the similar but steadily declining welfare from the wealth they get to allocate. Which brings us to our first result: If our single producer and our single consumer begin with equal endowments, and if the difference between consumer and producer surplus is not large, than the letting the market clear is likely to maximize welfare. But if our producer begins much wealthier than our consumer, enforcing a price ceiling may increase welfare. If it is our consumer who is wealthy, then the optimal result is a price floor. This result, a product of unassailably conventional economics, comports well with certain lay intuitions that economists sometimes ridicule. If workers are very poor, then perhaps a minimum wage (a price floor) improves welfare even of it does turn out to reduce the quantity of labor engaged. If landlords are typically wealthy, perhaps rent control (a price ceiling) is, in fact, optimal housing policy. Only in a world where the endowments of producers and those of consumers are equal is market-clearance incontrovertibly good policy. They greater the macro- inequality, the less persuasive the micro- case for letting the price mechanism do its work.

...

Let’s consider another common case about which many economists differ with views that might be characterized as “populist”. Suppose there is a limited, inelastic supply of road-lanes flowing onto the island of Manhattan. If access to roads is ungated, unpleasant evidence of shortage emerges. Thousands of people lose time in snarling, smoking, traffic jams. A frequently proposed solution to this problem is “congestion pricing”. Access to the bridges and tunnels crossing onto the island might be tolled, and the cost of the toll could be made to rise to the point where the number of vehicles willing to pay the price of entry was no more than what the lanes can fluidly accommodate. The case for price-rationing of an inelastically supplied good is very strong under two assumptions: 1) that people have diverse needs and preferences related to the individual circumstances of their lives; and 2) willingness to pay is a good measure of the relative strength of those needs and values. Under these assumptions, the virtue of congestion pricing is clear. People who most need to make the trip into Manhattan quickly, those who most value a quick journey, will pay for it. Those who don’t really need the trip or don’t mind waiting will skip the journey, or delay it until the price of the journey is cheap. When willingness to pay is a good measure of contribution to welfare, price rationing ensures that those more willing to pay travel in preference to those less willing, maximizing welfare.

Unfortunately, willingness to pay cannot be taken as a reasonable proxy for contribution to welfare if similar individuals face the choice with very different endowments. Congestion pricing is a reasonable candidate for near-optimal policy in a world where consumers are roughly equal in wealth and income. The more unequal the population of consumers, the weaker the case for price rationing. Schemes like congestion pricing become impossibly dumb in a world where a poor person might be rationed out of a life-saving trip to the hospital by a millionaire on a joy ride.

It's a very reasonable question to ask, especially in today's age, when income inequality is suspected to be as high or higher than ever. Perhaps even more so in the Bay Area, where income inequality is on par with that of developing nations.

And in such conditions, market clearing may not maximize welfare. As Matthew Yglesias notes:

When people say that a price-based scheme for rationing water is most efficient, they mean that prices will deliver the most efficient distribution of dollars and water. The idea is that how much people are willing to spend on something is a good proxy for how much they care about it, or how important it is to their well-being. Different people like different things, but you can buy all kinds of different stuff with dollars, and seeing what people choose to spend their money on tells you a lot about their preferences.

But dollars aren't a perfect proxy for well-being, because money means different things depending on how rich or poor you are. To a middle class American, $5,000 is a really big deal. To a multi-millionaire like Mitt Romney or Hillary Clinton, it's totally trivial — the value of their stock portfolios bounces up and down by that much all the time. To a person living paycheck-to-paycheck with no access to credit beyond very expensive payday loans, $5,000 could be a life-changing amount.

The technical term here is the "declining marginal utility of money." A given dollar produces less happiness in the pockets of a rich person than a poor one. That means that in a society with substantial economic inequality, an efficient distribution of dollars and water isn't going to be the same as an efficient distribution of happiness and water. This is what we're seeing in the North Carolina water case — the dollars are just a lot more important to the poor than the rich, so all the burden of adjusting to reduced water usage falls on them.

The reflexive response when it comes to many folks in tech on issues like Uber's surge pricing and net neutrality is to bow down before the power of the free market, and I count myself generally in that camp. Generally, it is an optimal scheme for efficient allocation of scarce resources.

I'm also generally sympathetic to companies pricing their products according the market. Apple charges a hell of a lot more for its phones than it costs them to manufacture them, but they've earned that surplus by producing a great product that people want. Uber charges surge pricing during certain times and enough people are willing to pay the multiple to get a ride because of the sheer convenience of the experience.

However, let's not blindly accept that it's welfare-maximizing for society without a skeptical analysis. As Waldman notes:

And there are lots of choices besides “whatever price the market bears” and allocation by waiting in line all day. Ration coupons, for example, are issued during wartime precisely because the welfare cost of letting the rich bid up prices while the poor starve are too obvious to be ignored. Under sufficiently high levels of inequality, rationing scarce goods by lottery may be superior in welfare terms to market allocation.