Here’s a striking figure: From 2012 to 2014, housing prices rose 13 times faster than wages.

That's the word from a new study by RealtyTrac, which looked at 184 metro areas around the country. One hundred and forty of those metro areas — with a combined population of 176 million — saw housing outpace wages. And 45 of them — for a combined population of 63 million — saw median home prices spike past 28 percent of median income for monthly mortgage payments. That's the threshold beyond which housing is traditionally considered unaffordable.

Overall, the study found that median wages went up 1.3 percent, while housing prices went up 17 percent over roughly the same time period.

This information opens a window on two things: why relative inequality is bad, and why inequality is likely occurring in the first place.

On the first point: Daren Blomquist, the vice-president at RealtyTrac and the author of the report, told Bloomberg Business that the housing recovery has "largely been driven over the last two years by buyers who are not as constrained by incomes — namely the institutional investors coming in and buying up properties as rentals, and international buyers coming in and buying, often with cash." For regular homebuyers to get a foothold, "either wages are going to need to go up or prices are going to need to at least flatten out and wait for wages to catch up."

One of the responses you often hear to concerns about inequality is that the purely relative gap between the bottom and the top doesn't matter. The only thing that counts is the material well-being of the bottom in an absolute sense. In other words, it doesn't matter that the "haves" own an increasingly large share of, well, everything because the "have-nots" can afford refrigerators and modern plumbing. But what we're seeing in housing is that, when you have a pool of relatively rich buyers, and a pool of relatively poorer buyers, with a big gap in between and relatively few to fill it, then prices will chase the rich buyers. That will drive the costs up for everyone. So the relative gap itself affects the absolute material well-being of the bottom.

From 1972 to 2005, the median family's spending on mortgages increased 76 percent. Similar increases happened for health care, cars and maintenance, childcare, college, etc. The price-chasing phenomenon is going on in goods and services across the economy. A typical family with two incomes in the early 2000s actually spent more of its budget on those sorts of fixed and universal needs than a typical single-earner family did in the 1970s.

But there's also something special about housing — or, more specifically, the land it's sitting on.

Last year, Thomas Piketty's Capital in the 21st Century diagnosed the accumulation of capital as the key driver of inequality across the world. As the value of capital increases relative to the value of the economy as a whole, the returns to that capital will displace returns to labor as a share of national income. A permanent upward ratchet effect sets in, driving inequality ever upward.

But Piketty's logic only holds if it's pretty easy for employers to substitute capital for labor — replacing workers with automation and so forth — and economists fiercely debate whether that substitutability is as high as Piketty assumes. Most capital also depreciates over time; buildings, cars, robots, computers, and other physical assets wear out and lose value. They have to be replaced. And presumably, the more capital you add, the less value each additional unit brings.

These forces should all combine to actually drive capital's share of income down over time. And Matt Rognlie, a 26-year-old economics graduate, just made a splash with a paper arguing it's done exactly that over the last few decades — if you don't include housing in your definition of capital. If you do include it, you get Piketty's ratchet.

And it's not so much the buildings themselves, it's the value of the land they're sitting on. Even more specifically, it's primarily urban land. Lastly, the value of housing and land is arguably the only reason the upper-middle class is able to keep pace with the truly wealthy in terms of capital accumulation.

Land is a pretty unique form of capital because its supply is largely fixed. If its value goes up, that doesn't necessarily mean you have any more land, or are getting any more productive value out of it. You can settle new land, but there's only so much useful land out there. And the most valuable land is valuable because it provides access to vibrant economies — i.e. it's urban, and close to already-settled land. Lastly, land and housing are often used as stores of value, which means the question of substitutability between capital and labor is largely moot.

That's why Peter Orszag recently argued that, instead Piketty's tax on wealth, we should aim for a tax on the value of land.

So was Piketty wrong to include land and housing in his definition of capital? No. Piketty wasn't trying to just explain economic mechanics, he was trying to capture social and political realities. As Thomas Hobbes and Adam Smith observed, wealth is power — the power to shape society and politics, and to shape the legal norms that undergird the economy's ability to exist at all.

You might say we shouldn't be focusing on inequality, but on distortions in the tax code — especially the mortgage interest deduction — that massively subsidize the value of housing for the wealthy. But that's a distinction without a difference. These tax distortions exist and are so difficult to kill precisely because they serve the powerful, and perpetuate inequality.

One of the conceits of the post-industrial internet age was that it would render the importance of land moot. But it looks like land and housing might largely escape the forces that drive down the value of most other forms of capital. Maybe they're the short-circuit in the economy that creates Piketty's inequality ratchet.