By Bob Berg and Jennifer Coskren, RISI
The U.S. housing market has been on a wild and protracted ride. Government policy and financial hubris in the middle of the last decade drove housing production to levels well above what fundamental demand levels would dictate and the market became over-bought (households financially ill-prepared to own a home were pulled into the market) and over-built (speculative). The Great Recession that began in 2009 brought the housing market crashing to well below underlying demand. In the aftermath of the collapse in the housing market, tight financing, a purged inventory of building lots and the relocation of labor constrained shelter production at levels well below those required to meet the recovering housing demand. Initially, this imbalance was masked by the large inventory of vacant units that resulted from the housing crash. At this point, however, the excess vacant units have been worked off (the inventory correction has been completed) and recovering demand is driving up home prices and rents as a result of the muted recovery in shelter production.
It is easy to dismiss the recent escalation in shelter prices as another "bubble." After all, the recent escalation in home prices and rents has pushed them up to, and in many cases above, levels that we experienced at the height of the boom years of the last decade. But the escalation in prices is the only thing these two cycles share. In fact, the factors leading to higher prices in these two "boom" periods were completely different and the ultimate market correction will be different.
In understanding the differences between the housing market in the last decade and the current market, it is important to understand that the key drivers for shelter demand and supply are underlying population growth and demographics, and that these factors are not impacted by economic cycles in the short run. As a result, cyclical swings in household formations are anchored by the underlying growth trends. Through the cycles, housing demand and production will cycle around the underlying trend dictated by demographics. Currently, population growth will require an average of 1.13 million units per year between 2014 and 2020 and demand for second homes and housing stock replacement will require another 0.34 million units per year (Table 1). This puts trend housing demand at around 1.47 million per year.
In the housing boom of the last decade, housing production outpaced the fundamental demand for housing and ultimately household formations collapsed as demand was exhausted and job losses mounted. There were no more marginal homebuyers to tap into and speculators were frozen by the buildup of vacant inventory and falling home prices. Housing demand could not be sustained at levels well above underlying demand. The correction for this imbalance was quick and brutal as demand and prices plummeted.
We are currently at the other extreme in the market where housing production has been running well below fundamental demand levels. Over the period 2009-2015, household formations averaged 0.848 million per year, while net additions to housing stock averaged just 0.577 million per year for a 0.271 million average shortfall in supply over the six-year period. Initially, this shortfall was masked by the large number of vacant shelter units on the market. In 2009, there were around 1.5 million vacant homes for sale and rent and banks repossessed nearly 1.0 million homes through foreclosure, according to RealtyTrac. However, the modest level of household formations in recent years worked through the inventory of vacant shelter, and by the middle of 2014 the total shelter occupancy rate had returned to pre-recession levels and apartment vacancy rates dropped to historical lows (Figures 1 and 2).
In the last decade, demand and production levels in excess of underlying fundamental demand led to a pool of vacant shelter which had to be cleared before the market corrected. During the recession, households failed to form mostly due to virtually nonexistent job growth as U.S. payrolls were slashed. And even as the recovering in the overall economy took hold, households continued to lag behind underlying demographics for reasons ranging from the prohibitive cost of shelter to high student debt. Those individuals that would have formed households have not gone away and they add to the pool of individuals that would form households under improved conditions (pent-up demand). Household formations did surge to levels above underlying demand and the pool of pent-up demand was drawn down, but in early 2016 it still stands at over 2.0 million. Household demand would move even higher and pent-up demand lower if the supply constraints hindering the market lifted. This pool of demand will only grow as long as production falls short of underlying demand.
In previous cycles (pre-2000), housing production quickly rebounded to levels in line with or above demand and the pool of pent-up demand built up during the downturn was quickly worked off. The recovery in housing in the current cycle has been constrained by the destruction of the physical (available land and labor) and financial supply chain. So, it was much easier to stop building in the last cycle to allow for the market to correct than it has been to restart shelter production to rebalance shelter demand and supply.
Resolution to the current housing crisis or supply shortfall will come from both market forces and public policy. Market forces are initiated through higher prices and increased investment. Higher prices will slow household formations as marginal entrants into the shelter market are forced to postpone their plans. Local escalation in prices should redirect corporate and small business growth plans to areas with more affordable housing (think San Francisco, New York City, and Boston shifting to Austin and Cleveland). On the supply side, higher prices will help stimulate investment in shelter. Shelter production doubled between 2009 and 2015 with multifamily housing accounting for nearly 50% of the growth as access to mortgage financing limited entry to the single-family market. On the face of it, this is a solid recovery; however, current levels of housing production are nearly 20% below the levels required to meet fundamental demographic demand.
Housing production needs to reach close to 1.5 million units just to meet underlying demand over the next few years, and this level does not allow for any reduction in pent-up demand. However, the extreme and prolonged downturn in housing has destroyed much of the supply chain in housing production. Building lots were purged from builders' balance sheets and a large portion of the labor either left the work force or migrated to other sectors. And the housing finance supply chain (builder and buyer) has yet to fully recover. The relatively slow pace at which the housing supply chain is being rebuilt provides the underpinnings for the current housing crisis.
The other part of the solution to the current housing crisis will come from public policy. The policy prescriptions, however, have to be directed at supply and not traditional demand remedies. For instance, restrictive building codes, protracted permitting processes, and strict zoning will have to become more growth oriented. Labor retraining programs and relocation tax incentives would also help. And, of course, accommodative financing for builders. Policies directed at increasing demand or controlling prices will only aggravate the current housing market conditions.
In either case, the housing market will not be back in equilibrium until production returns to levels that will allow for the elimination of pent-up demand. There is no "new reality" in housing market. Ultimately shelter production will have to return to levels that will accomplish this rebalancing, and market and policy pressures will work toward this solution. And unlike the previous boom, this correction has already and will continue to be a drawn out process.