Local Zoning and Hukou – Systems of Mobility Management in the US and China

The writer Michael Hobbes had an excellent thread on housing and homelessness the other day:

As Hobbes continues to say, “Every American city that sees a boom in incomes and employment also sees a spike in homelessness. Our deranged housing policy has made equitable urban growth a thing of the past.”

The relationship is indeed causal. The poster children for the current housing crisis in America are precisely those cities where high amenity construction and an influx of college-educated workers has not been followed by sufficient homebuilding – San Francisco, Boston, New York and Austin; to a lesser extent Los Angeles and Seattle too. With no other alternatives, these new high-income workers bid up the existing housing stock – this is the gentrification process – and research by Zillow has found that future increases in homelessness is well-predicted by what comes next, i.e, increases in average rent. While the natural solution to an increase in the demand for housing in a place is to increase the supply for housing in that place, the issue has instead festered and spawned a kaleidoscope of groups all pushing their own solutions, each with a worse tradeoff than the last.

In a classic case of strange bedfellows, some NIMBYs and pro-homeless activists have pointed out that in fact there is no housing crisis, since the national housing stock is more than capable of housing the unhoused, with there being more vacant homes than homeless people. This is true, but subject to an obvious critique: since extra houses in Detroit can’t be moved to Culver City, the only option remaining is to move some of the people out of the crowded cities into these extra homes. And while there are thoughtful approaches to this (like pursuing geographic redistribution as a tool of federal policy), more often than not proponents of this approach end up advocating for a deep and modern form of segregation.

As it happens, the mismatch of homes and jobs is not a uniquely American problem. China, amid its stomach churning rise to middle-income, has been groaning under the weight of a similar problem for decades. I want to compare the restrictions on mobility imposed by local control of zoning in the US with the top-down system used in China – the hukou, which is a form of census and identification strategy that maps each Chinese citizen to a region and determines their eligibility for legal residence in other regions. Obviously, hukou is a rather more domineering way to control the movement of a population. But the effects of local control of zoning are no less severe; rather remarkably, the natural whims of homeowners in Marin County and the Upper West Side have come to align with the technocrats at the CCP.

“Getting to Denmark”

First some accounting: every population on Earth has become more urban since the Industrial Revolution, but the Americans (beginning in the 1850s) and the Chinese (beginning in the 1970s) were really good at it.

It was only by 1920 that the US would become a majority-urban nation; today about 80% of Americans live in cities. The late 20th C. suburbanization is evident to see in the below chart as a break in trend, as is the great return of the millennials to cities after 2008; to a first approximation, however, it’s easy to see that urbanization followed a linear trend up for several decades before slowing at higher levels.

The Chinese have taken a much more volatile path to urbanization.

In 1949, the Communist Party took power in a country that counted 60 million urban residents, a mere spearhead compared to the 450 million living in the countryside.

Early on, they established the hukou, or “household registration” system, a tool of the command economy designed to help the party manage the vast Chinese population. Hukou were used to designate where residents would receive government benefits, like healthcare, schooling for their kids, and work eligibility. They came in two flavors, rural and urban, and while changes to hukou status were nominally allowed, in practice they were exceedingly rare: maybe something like 1% of all rural holders could have converted to urban status each year.

China Loosens Hukou Residency System to Spur Growth - Bloomberg
These are hukou – they’re physical books, sort of like IDs, with all the requisite data about each individual on them.

Hukou shaped Mao-era policymaking. As the Dear Leader’s focus swung from rural collectivization to leap-frog industrialization and then back to iconoclastic civil struggle in the late 1960s, holding an urban hukou was tantamount to a stay of execution – 19 in 20 deaths during the Great Chinese Famine were of rural hukou holders, as they were expected to feed themselves with the surplus remaining after CCP authorities had collected their grain targets, while urban residents received (meager, but reliable) state rations.

Even after the death of Mao and up through the 80s, modern megalopoli like the Pearl River Delta were agricultural, still covered in “rice paddies and water buffalo,” as one urban studies scholar puts it. Shenzhen changed all that. Amid the raft of economic reforms which opened up special economic zones from 1979 onwards was a crack in the armor of hukou that allowed for rural merchants to legally bring their business to cities distinct from their own.

Further reforms in 1992 would open more benefits and a form of legal residency to migrant laborers in a few cities; yet more reforms in the early aughts would finally tear down the rural-urban distinction, replacing it instead with a “regional” hukou. Those reforms have basically worked:

Statistic: Degree of urbanization in China from 1980 to 2020 | Statista

By the mid-2000s, poor farmers in the Chongqing hinterlands had it well within their rights to cast down their plows and pick up factory tools – but only in Chongqing. They couldn’t leave behind the farm as easily for one of the dynamos like Shenzhen, Shanghai, or Beijing. And therein lay the rub.

Between countries, within countries

While overall urbanization stats tell a simple story, they hide the changes in mix between cities which has attained a greater salience today. Put another way, like the UN does in 75-year report on Inequality, “Income inequality between countries has improved, yet income inequality within countries has become worse.”

The below chart illustrates this well – taking the per capita personal income of the top 100 combined statistical areas in the US, we observe that the playing field has become much less equal. In 1969, the median CSA had personal income only 25% lower than the peak, in that year oddly enough represented by Reno, NV. In 2019, the median CSA had a personal income nearly 50% lower than the Bay Area’s.

Source: Bureau of Economic Analysis

This development has important implications for individual behavior and the study of urbanization more broadly. If, say, rural productivity in 1969 was equal to an index of 25 on the chart, then even someone moving to the poorest CSA at that time – New Orleans, LA – could have expected to nearly triple their personal income.

By contrast, say rural productivity was again at 25 for 2019, then someone moving to the 100th-richest CSA, Lexington KY, would only expect to double their income. That 2019 mover has to get all the way to Minneapolis to triple their income like the person who moved to the Big Easy in 1969.

Put simply, while the returns to urbanization were rather uniform in the mid-century, the distribution has skewed in our day. All urbanization is no longer created equal and the restrictions which prevent workers from moving to San Jose and Seattle and NYC and Boston are a club to the kneecap of broader American growth.

Two clever economists advanced this argument a few years ago and did the legwork too; they found that the increase in GDP from ending restrictive housing regulations and allowing greater mobility was astronomical – from 1964-2009, the total increase in GDP would’ve been +36% higher than it was in reality, worth an extra $3.4tn in 09 dollars (Hsieh & Moretti, the authors of that study, corrected an error caught by the economist Bryan Caplan, which had actually underestimated the theorized effects by something like $2tn dollars!).

Of course, for all this to work you have to believe that the increase in income inequality between cities has actually come along with a decrease in mobility to those rich cities. But that does seem to be the case – as above, we noted that these richest cities are ground zero for the housing crisis, places where housing costs are so high not only as to deter all the potential newcomers who don’t have FAANG RSUs as part of their compensation, but also to actually evict present residents!

The real economists have weighed in on this for a while too – the impacts of lower mobility are at the core of Tyler Cowen’s Complacent Class. Davis & Haltiwanger (2014) found rates of job relocation fell by a quarter going into the 90s, while Molloy et al. (2011) declared internal migration to be at a 30-year low.

Molloy et al. (2011), p. 174

If that’s not enough for you, Brookings laid it out cleanly in a 2018 note called “Americans aren’t moving to economic opportunity” – more than half of migration out of “low vitality” counties was destined for other “low vitality” counties, and only 13% of migration from lowest vitality counties made it to highest vitality counties.

The Chinese economy has been no less susceptible to this phenomenon. The mid-aughts hukou reforms which got rid of the urban/rural split brought in new dividing lines, this time in the form of regional boundaries. This has caused its own chafing issues as poor rural residents want the gains from income possible in Shenzhen & Beijing and impossible in their local tier 3 urban area. Filmmakers like Jia Zhangke, who have made it out to the coastal superstar cities, pay testament to these inequities – the lives of a group of miners in remote Shanxi province are depicted in A Touch of Sin as being all sound and fury, bitter cold and hard labor for meager wages.

It’s no wonder that, as The Economist found in 2020, the smallest Chinese cities are actually seeing their hukou-compliant populations fall. Rural hukou holders have a certain “reservation” status, similar to job-seekers in a labor market, below which they will prefer to retain their rural status and forgo the social benefits of registration in the city in which they work. Xi Jinping, who wrote as a student about tinkering with hukou, has pushed a new initiative to spruce up these smaller towns and cities, with little word back on success.

Hukou today, as The Economist has it, is a game of three tiers: you’d love to but can’t get into Shanghai and Beijing, you’re ok with and can get into Chengdu and Xi’an, and you don’t even want to get into small cities like Dawan or Wuwei.

NIMBYism on the Pacific Rim

Nowadays the arguments for keeping strong hukou distinctions are weaker than ever. Faced with four years of Trumpian jousting, COVID-19, and a burgeoning demographic crisis, the party under Xi has been more and more active in tinkering with the rules and allowing for more migration, a strategy once referred to as adding a “new engine for the slowing economy”.

Where opponents of liberalization usually stake their claims is in arguments about the cost of providing for all those newcomers. In a very telling discussion, Fei-Ling Wang, a professor at Georgia Tech, describes the modern pressures on the reform camp:

Resistance to any proposed major reform now, ironically, comes from the urban people – not ironically, but expectedly I guess – and not necessarily from government planners. Government planners are hoping to speed up urbanization even in large cities so the economy can get a new driving force. However, urban officials, the local cadres, see this as compromising their current privilege. They see the rural folks as coming in to share their fruits. The urban-rural difference in China is horrendous – living standard, income, education, medical service, you name it. This great gap really makes the urban folks feel they are privileged and wonder why they should share the scarce resources with country folk. The local officials also worry about it [hukou reform], as it may get local people really angry or create a management issue.

Fei-Ling Wang, quoted in The Diplomat

There we have it – Chinese NIMBYism made flesh. What in the end distinguishes the ideas of those who were there first, either in China or the US, from those who would like only the chance to share in the fruits of high-productivity economies? Very little, except for a commitment to keeping the rest of us all a little poorer.

Re: The NYT Piece on Extended Stay America and Modern SROs

Mya Frazier has a blockbuster of a piece out in the NYT Magazine covering the people who live in long-term hotels, a group one of her sources refers to as the “precariously housed”. It’s a very well-balanced read, taking the proprietor of one chain, Extended Stay America, to task for outsized room fees and brutish eviction practices while also lambasting high growth regions like Columbus and Atlanta for blocking the building of new housing. Taken as a whole, this piece (including the asides about revolting check-cashing fees) presents a crystal-clear accounting of the “high costs of being poor”.

I wrote on Twitter that I see the key culprit here as being lack of new housing, which I am wont to do. The piece also highlights the decline in single-room occupancy hotels, which were once a robust part of the lowest-income housing market but were regulated away to death. But for her part, Frazier focuses more on the role of credit scores in denying many poor Americans access to the rental market.

The criticisms are all spot-on here, including a nice piece of grapeshot fired at the chancy figure of Richard Cordray, who in 2012 became the first boss of the Consumer Financial Protection Bureau after Elizabeth Warren decided to forgo leading the agency she’d molded and remain in the Senate. Cordray left the agency to go lose state-wide for the second time to Mike DeWine in the 2018 Ohio gubernatorial race but apparently landed on his feet in Biden’s Dept of Education, running student loan claims.

In her piece, Frazier describes how the former director helped the credit-report triumvir of Experian, TransUnion, and Equifax pivot their businesses after the housing crash to peer deeper into the lives of poor people. They began to include “‘alternative data,’ such as rent and utility payments, into credit reporting…Cordray saw the push as good public policy — a form of ‘financial empowerment.'”

Credit reports have a shallow history – they were only used for the first time in 1989 – but as everyone knows, they control more than just prices for credit, the interest rate borrowers face; for people with bad credit scores, their score can lock them out of credit entirely or be used as a proxy rationale for institutions to enter into non-credit transactions with them.

Case in point: the usage by landlords of credit scores for offering apartments. There is no prima facie reason why a landlord should be able to deny you an apartment in February because they worry about your payments in October – if October comes and you can’t pay, you leave and it’s on the landlord to find a new tenant.

It’s kookier still to believe that that February rent should depend on credit mishaps from years prior. But that’s how the industry has developed – landlords have been able to impose on renters everywhere the 12-month lease as a technology of brutal repression, making a simple payment for services into a loan-like contract.

(One line of counterargument might emphasize that there are benefits to renters for this long-term contract, but I’m skeptical: when times are good, you can probably get a raise, so you could afford higher rent; but when times are bad, you lose everything, so you can’t afford the prior level of rent anyway. Paradoxically, longer maturities benefit the creditor here.)

This point leads to two further and alternate observations. The first is that we should abolish credit scoring. This is seductive, but wrong-headed, because as the Frazier piece points out, credit scores were a technology that facilitated broader access to borrowing: “Credit scores…stratified credit markets. If you had a low score, you paid higher interest rates; once the process was automated, decisions that used to take days or weeks were made within hours.” That improvement in lending decision speed was useful, and shouldn’t be thrown away.

The second observation is on sounder footing: if we must be stuck with a minority precariat, long-term leases, and predatory credit reporters, then credit scores must be reformed to move faster. If you get a new steady job, the weight of your prior credit should fall sharply. Entries in your credit history should stick around for a couple of years at most. Landlords should be steeply disincentivized against declining because of credit score, and instead should be biased towards green-lighting applicants unless something is incredibly wrong.

Siegel Suites, the business operating the hotels where the piece is set, has carved a middle way through this dilemma by ignoring credit histories and offering short-term rentals. This is a good thing: for lack of Siegel, the women profiled in Frazier’s piece would be flat-out homeless, as indeed her first subject was until she and her father learned about Siegel’s offering.

However, by being the only game in town willing to take on these higher risk renters, Siegel is able to charge exorbitant premiums above market rates. Of her second subject’s budget, Frazier writes, “For about five months, Green and her partner managed to cover the initial daily room rate of $44.99, which would add up to more than $1,300 a month — almost twice the fair-market rent of $717 for an efficiency apartment in Columbus, as calculated by HUD.” Renters at Siegel Suites, started by a one-time NoHo body-shop owner, and the better-known Extended Stay America, deserve more competition in this space, which would drive down these painful premiums to rational levels. These landlords deserve some extra profit from the risks they’re taking, but not very much.

(As an aside, I’ve been curious about the model for the landlord decision to evict, prompted first by the closure of the Cinerama theater. When is it more advantageous to evict and reenter the tenant search market, rather than restructuring payments in arrears [even at 100% haircuts] and allow the current tenant to return to normal payment schedule? The fact that it took a CDC order to prevent evictions during the pandemic suggests landlords are quick to evict, a tendency too strong to be fully rational.)

I criticized a recent story from a trio of NPR reporters which tried to account for the large racial homeownership gap by means of credit score discrimination, because I really don’t believe that it’s responsible for driving much of the variation in the purchasing process. But I am much more inclined to believe Frazier’s account, that recourse to the three-digit number in rental markets has gone too far.

At the end of the day, you need a house over your head at whatever price you can afford, and it boggles the mind that the market for housing is insufficiently thick to provide that. More public investment is likely the answer.