Commentary: Before Atlanta even thinks about winning Amazon’s HQ2, it needs to get serious about housing
Commentary: Amazon wouldn’t cause a housing shortage for Atlanta. We’ve done that to ourselves.
Atlanta has a strong chance of landing Amazon’s second North American headquarters, nicknamed HQ2. We routinely place in the top three cities among experts and observers, and the possibility is the topic I get asked about the most when people learn I work in affordable housing. The prospect of 50,000 high-paying jobs—the average salary would be $100,000—at Amazon’s HQ2 is both exciting and concerning for Atlantans, particularly Atlantans already anxious about rising housing prices.
Measuring Amazon’s impact
This concern is understandable: residents see an out-of-control housing market in Amazon’s home base of Seattle, where rents are increasing 7 percent per year and home prices are up 12 percent per year, and they rightfully worry about what Amazon would mean for Atlanta. To help understand the scope of Amazon HQ2, let’s compare it to Midtown, Atlanta. Approximately 54,000 jobs are located between North Avenue and the Brookwood interchange, and only 30,000 of those make more than $40,000 per year. Amazon’s HQ2 would be the equivalent of another Midtown, and then some.
If Amazon does choose Atlanta as the home of HQ2, the company won’t place all 50,000 of those jobs next year. Its proposal says those jobs will be spread out over as many as 17 years. If those jobs were spread out over 16 years, that’s 3,125 jobs per year. By comparison, metro Atlanta has added between 60,000 and 90,000 jobs every year since 2013. Amazon is adding a lot of jobs, but at a level that is also well within an annual variation for metro Atlanta.
The potential impact of Amazon-in-Atlanta on housing, however, is measurable. Not all of those jobs would be filled by people who have to move to the region, but for example, let’s say they were. Metro Atlanta currently has 0.8 households for every job in the region, so under this scenario we could reasonably expect Amazon to create a need for 2,575 new homes per year. By comparison, metro Atlanta permitted 35,000 new housing units in 2016 and was on pace for about 33,000 units in 2017. The metro region would need to build 7.6 percent more homes than we are building right now to house all the new jobs that Amazon would bring to the city if every employee were moving to the region.
Can we handle that many jobs and new households?
Whether that number sounds preposterous or easily achievable to you probably depends on your perspective. If you live near one of the region’s booming multifamily markets, where it seems like every other week a new high-rise or mid-rise apartment building is breaking ground, you probably think our current level of construction is unsustainable. But what if I told you that for more than twenty years, from 1985 to 2007, we permitted more than 50,000 new homes per year to be built in the metro region? From 2000 to 2006 we even permitted 67,000 new homes per year. That’s more than enough to house all the new Amazon workers.
You may not believe it from all the cranes you see, but we still haven’t recovered from the housing crisis. From 2010 to 2016, we’ve only built two-thirds of the housing we need. We have built half of the number of single-family homes that we built from 1990 to 2010, and the biggest gap in housing production is in affordable, entry-level homes. The supply of existing homes for sale is still depressed. With few options available to purchase, these would-be-homeowners are instead continuing to rent, keeping apartment vacancies low and rents high.
Atlanta isn’t alone; this is a national trend, although it is worst in markets with lots of development restrictions, such as New York, Seattle, and San Francisco. Housing prices are rising in Atlanta and elsewhere because we simply have not been able to meet the demand for new housing. Even if we were closer to meeting demand now, in 2017, we still have half a decade to make up for. If the city is to add as many homes as city planning commissioner Tim Keane thinks we can add—900,000 additional residents by 2050—we will need to build twice as many new housing units without any loss of existing housing for 34 years straight. If we want to see a meaningful impact in metrowide housing affordability, we need to at least double housing production from where we are today.
We are in the midst of a near unprecedented housing shortage, whether we get Amazon or not. Everyone suffers when we don’t build enough homes. The most affluent households bid up prices for the most desirable homes, which drives prices up for everyone. This shortage hits the most vulnerable first and hardest—there has been a lot of focus on how the BeltLine has driven up local housing prices, but that impact wouldn’t be as dramatic if we didn’t have a metro-wide shortage. Middle-income households are struggling with increasing housing prices as well. Eventually our housing crisis will start to limit the region’s ability to support new jobs and economic activity.
This trend is recent and is the result of policy choices.
This is not an unsolvable problem. Metro Atlanta home prices were essentially the same for 20 years. From 1977 until 1997, inflation-adjusted home prices went up 3.2 percent overall, which is less than prices have risen in 2017 alone. We achieved this price stability by building enough low-cost suburban housing to meet our growing demand. We started having problems when we began to reach the limit of suburban-oriented development and Americans began to return to cities, where legacy land uses and zoning restrictions limit new development.
We have to recognize that our increasing housing values are directly related to how much new housing we are—or more specifically are not—building. It is not an accident that real home prices in metro areas with lots of restrictions are up by anywhere between 150 percent and 300 percent since 1977. Meanwhile, prices in Atlanta, Dallas, and Charlotte are up only 20 percent to 30 percent. Amazon’s explosive growth has had an impact on Seattle’s housing market because Seattle was not prepared to deal with such an increase in demand.
How can we begin to solve metro Atlanta’s housing crisis? We can’t single-handedly fix the credit markets and consumer-saving patterns that have limited the construction of entry-level, single family homes. But we can start identifying barriers to new construction overall. Policies such as inclusionary zoning, which address only pieces of the problem, are helpful, but insufficient to the scope of the issue. Every unit of housing makes a difference, regardless of whether it is a luxury unit or more affordable. We could double or triple the amount of affordable housing we build annually and it would not make a dent in the additional 20,000 to 30,000 units a year that the city needs to slow rent and price increases. Municipal-level solutions can make an impact, but the city of Atlanta alone can’t increase development enough to make up for a 20-year, metro-wide housing shortage. Every level of government needs to evaluate what it can do to create more housing.
We have to fix the underlying problems: zoning and land sale markets.
The first thing we have to do is tackle the broken land market we have created through decades of poor zoning and incentive granting, which keeps too much developable land off the market and limits development to parties who know how to work the system. When nearly every development of scale requires either rezoning or subsidies in order to be built, we have to admit that the basic model is broken. Zoning which is inadequate to current market conditions incentivizes landowners to wait for the most aggressive offer—this behavior limits supply, drives up prices, and ultimately results in fewer new housing units. The housing that does get built is the most dense and most expensive possible, and it is that much more expensive to subsidize low-income housing. Our current zoning system adds risk, time, and cost to development, but the worst impact is what its unpredictability does to the availability and price of developable land.
We need a form-based code that allows enough density for the next 40 years of growth in areas that can actually support new construction, combined with a land value tax that moves vacant and underutilized land into the market for development. The use of development incentives by local governments for market-rate development—which in the end most often benefit landowners and perversely can make development harder—should also be dramatically scaled back by the state.
These are by no means the only policy changes that can help the housing shortage. Others include innovative modular construction methods, reducing the impact of parking on new construction through more walkable and transit-oriented development, and increased funding for affordable housing. These measures are all critical, but if they are not paired with policies which address the underlying restraints on our housing supply, they can at best only make a dent in our shortage. Policies which increase the cost of development could actually make things worse. The longer it takes to start building enough, the more catch-up we will have to do. The first step, however, is diagnosing the problem correctly.
Ben King is a development manager with an Atlanta-based developer, where he oversees the planning and construction of mixed-income, affordable housing in Atlanta and Dallas/Fort Worth, Texas. He has previously worked for private developers and municipalities across the country as a consultant on market feasibility, housing economics, and public-private finance.