Pages

Sunday, November 05, 2017

Urban Climate Change Adaptation and Local Real Estate Markets

For those who wonder if a Department Chairman can get some work done, here is the introduction of my new paper that I will present at the Hoover Institution on 11/8/2017.

Urban Climate Change Adaptation and Local Real Estate Markets

Matthew E. Kahn

University of Southern California and NBER


Introduction

The major productivity hubs in the United States are located in coastal areas such as San Francisco, Seattle , New York City and Boston (Hsieh and Moretti 2017).  In each of these areas, a set of high technology firms and high human capital workers have co-agglomerated creating highly productive clusters.   These cities both attract talent and the close physical proximity between these workers and firms causes better matching between workers and firms such that cross-firm learning takes place (Combes et. al 2012, Glaeser 1998,  Rosenthal and Strange 2004).
Such coastal productivity centers raises concerns that natural disaster risk and climate change will impose enormous costs for the U.S  because it could disrupt economic activity.  In early September 2017,  Hurricane Harvey shocked the Houston economy and Hurricane Irma significantly damaged Florida. These events highlight how natural disasters can impact real estate capital.  While the science of climate change features many open questions, we have an increased understanding that different geographic regions will face more extreme temperature and rainfall events and that tail risk of severe natural disasters could worsen (see https://www.ipcc.ch/report/ar5/wg1/).
The economic consequences of these geographic shocks hinges on how and where we build our cities.  Over decades, we have made durable investments in capital and infrastructure that place millions of people and billions of dollars of capital in areas that could be at increased risk of sea level rise and other challenges posed by climate change (Changnon et. al. 2000, Pielke et. al. 1998, 1999).
Zillow’s researchers have made scary predictions about the aggregate capital losses (perhaps $400 billion in Florida alone) that might occur in the year 2100.[1]  This prospective research overlays maps of current coastal assets with different scenarios of future sea level rise.   An emerging climate economics literature studies the historical relationship between geographic places (such as nations or counties) and examines how their economic growth and population growth co-varies with climate conditions (Hsiang 2016). This research has documented a negative correlation between average temperatures (i.e summer heat) and economic growth (Deryugina and Hsiang 2014).    
The Lucas Critique teaches us that past historical relationships may not yield good forecasting rules if economic decision makers reoptimize “as the rules of the game change” (Lucas 1976).  While Lucas originally focused on how individuals respond to changes in government policy, this same logic applies in thinking through how individuals respond to changes in climate patterns.
Starting with the early work on rational expectations, economists have emphasized that investment patterns are a function of future expectations (Lucas and Prescott 1971).  If investors “know that they do not know” the likelihood of fat tail risks, they will be less likely to make irreversible large sunk investments.  Such rational agents will instead seek a series of less costly investments that offer the option to wait and see how the threat a specific area faces (Bunten and Kahn 2016). 
Expectations of changing climate conditions drives investment patterns and these investments facilitate the creation of real estate structures, neighborhoods and cities that are more robust to the major challenges that climate change will pose.  For example, areas in Florida that flood during hurricanes will be more likely to face this same risk in the future.  Such geographic predictability creates the possibility for forward looking people and firms to invest in several strategies to cope with risks.  These include; where we live, the structures we live in and our investments in resilient infrastructure and disaster preparation.
This logic challenges the recent climate economics literature that implicitly assumes that economic agents do not form rational expectations and thus do not plan ahead based on anticipated trends.   There is a certain irony that the “new climate economics” implicitly embraces the 1950s model of cobweb expectations formation as economic decision makers base their future expectations solely using the past historical record.
This paper melds ideas based on the efficient markets hypothesis from finance and the rational expectations model from macroeconomics to generate a series of empirical hypotheses for testing whether our urban economy and real estate assets are increasingly resilient in the face of emerging climate risk.
While this paper is prospective, I present a set of testable claims for judging the role of free markets for facilitating adaptation at ever lower cost.    A central premise of this paper is that cities compete for jobs and the high skilled.  If a city such as San Francisco becomes riskier and less livable, then it could certainly experience a brain drain but urban people will substitute to their next best alternative.  If no alternative city is a close substitute for San Francisco, then those who greatly demand San Francisco’s amenities do lose the most.  Such loss in consumer surplus is a cost of climate change.   In his 2002 AEA Presidential Address, Sherwin Rosen (2002) wrote: “Markets accommodate diversity by establishing prices that tend to make different things relatively close substitutes at the margin.”
Some cities and neighborhoods may face extreme climate induced challenges, other geographic areas may have a comparative advantage in coping. If extreme heat is the challenge, then coastal cities at northern latitudes may have an edge.  China has demonstrated that multi-million person cities can quickly be built. The U.S is blessed with ample land.  If just ½ of Montana was built up to Hong Kong population density, 1.5 billion people could live there.  This example highlights how we might build our future cities if local land use regulation does not inhibit such construction (Kahn 2011, Kerkoff, Prescott, Ohanian 2017).
For geographic areas such as Phoenix that are predicted to face increased extreme heat due to climate change, financial economics makes strong predictions about how assets tied to Phoenix will be affected.  The theory of compensating differentials and the efficient markets hypothesis predicts that asset prices (i.e home prices) reflect all available information about the current and future expected dividends.  In the case of real estate markets, the dividend is the flow value of productivity and amenities enjoyed by living in a given location.   If an area’s productivity and amenity value is expected to decline in the near future because of climate change, then home prices today will be lower to reflect this future reduction in the dividend.[2]  If incumbent home owners are the residual claimants on “bad climate news”, then they become an interest group lobbying local officials to make investments that boost the city’s resilience (i.e infrastructure that is more flood resistant, electricity systems that can withstand heat and storms). Such individuals will demand adaptive products such as improved housing and better air conditioning.  
The demand for adaptation friendly products creates a market for entrepreneurs who can supply them (Acemoglu and Linn 2004, Kahn and Zhao 2017).  I will discuss the technological and entrepreneurial opportunities created by extreme climate conditions.  Whether the best of these products can offset much of the new risks that Mother Nature will expose us to remains an open question.  The ongoing decline in deaths from extreme heat observed in both the United States and in the developing world suggests grounds for optimism here (Baccera et. al. 2016, Burgess et. al. 2016).  When we anticipate a trend, we have the resources and the incentive to take actions to reduce our risk exposure.    The death count from natural disasters declines with economic development and the diffusion of information technology such as cell phones (Kahn 2005). 
While this paper embraces an optimistic view of our collective ability to adapt, some subgroups of the population face adjustment costs that could limit their set of feasible adaptation strategies.  Consider migration costs.  The elderly and the less educated are less likely to migrate. I explore the causes of such migration costs such as demographics and built up place based social networks.  For many people, their identity is tied to a specific place (Akerlof and Kranton 2000).  I discuss emerging technologies that are likely to attenuate these issues.  A final subgroup of the population who may face adaptation challenges are those with a propensity to under-estimate emerging risks.  This group will neither support mitigation strategies such as a carbon tax nor will they take pro-active costly steps to protect themselves. 
Local, state and federal laws set the rules of the game that affect the spatial equilibrium.   Building on past research on the unintended consequences of government policy, this paper will highlight several cases in which government rules impinge on adaptation.  Free market adaptation economics offers new insights about the costs of government activism (Kahn 2010).



[1] https://www.zillow.com/research/climate-change-underwater-homes-12890/
[2] For example of climate economics that quantify how places are affected by natural disasters see https://www.nytimes.com/2017/09/29/opinion/puerto-rico-hurricane-maria.html?utm_source=NewsClips+Subscribers&utm_campaign=6afeaca480-EMAIL_CAMPAIGN_2017_08_22&utm_medium=email&utm_term=0_6ac49021ff-6afeaca480-154294709&_r=1. This literature abstracts away from quantifying the cross-elasticity of where labor and capital flow to as the origin area is shocked.