May 2013 A Nation of Gamblers: Real Estate and American History By Edward L. Glaeser (Harvard University)

Introduction what are the policy implications of the Taubman Center Policy Briefs are short Great Convulsion? overviews of new and notable research Between 1996 and 2006, there was on key issues by scholars affi liated with a 53 percent real increase in housing This policy brief, drawn from a more the Center. This policy brief is based on detailed lecture and paper1, examines the text of the 2013 Ely Lecture presented prices in the . Between at the American Association 2006 and 2011, real housing values America’s long history of real estate convention in San Diego. The text of the speculation to shed light on recent lecture is available in full in the forthcoming decreased by 28 percent. This sharp issue of the American Economic Review. increase then almost equally sharp events. First, rural land speculation is Edward L. Glaeser reversion to the mean was the Great examined across three time periods Edward L. Glaeser, the Fred and Eleanor Housing Convulsion. —18th Century Speculation in New Glimp Professor of Economics in Harvard University’s Faculty of Arts and Sciences, York State, the 1815-1819 Huntsville Figure 1 shows how the increase, then became the Institute’s director in June 2004. Cotton Land Boom, and the 1900- Glaeser, who also directs the Taubman decrease, in real estate prices was Center for State and Local Government at 1940 Iowa Wheat Land Cycle. Next, Harvard’s Kennedy School of Government, consistent across the country but the three episodes of urban speculation, is the author of numerous papers on the magnitude varied widely. The fi gure determinants of city growth, the role of confi ned to specifi c cities, are studied cities as centers of idea transmission, and a shows the relationship between the variety of other urban-related topics. Glaeser – The Chicago Boom: 1830-1841, Los change in the logarithm of Federal is the author of Triumph of the City: How Angeles in the 1880s, and New York Our Greatest Invention Makes Us Richer, Housing Finance Agency (FHFA) Smarter, Greener, Healthier, and Happier City: 1890-1933. Last, metropolitan (Penguin, 2011); editor of the Quarterly prices between 2000 and 2006, and speculation, spread across urban Journal of Economics; the co-author (with the change in the logarithm of FHFA Alberto Alesina) of Fighting Poverty in the areas and their surrounding suburbs, US and Europe (Oxford University Press, prices between 2006 and 2012. It 2004); the co-editor (with Claudia Golden) is reviewed across three periods and illustrates both the remarkable amount of Corruption and Reform: Lessons From locations – “The Housing Bubble America’s (University of mean reversion (the regression of Chicago Press, 2006); and the editor That Didn’t Happen” after World of The Governance of Not-For-Profi t coeffi cient is -.85) across areas and the War II, California in the 1970s and Organizations (University of Chicago Press, tremendous heterogeneity across the 2003). He received a Ph.D. in economics from 1980s, and, fi nally, the Great Housing the University of Chicago. United States. Convulsion: 1996 – 2012. From these Economists have now studied nine episodes, six key lessons are this Great Housing Convulsion drawn. extensively, but many questions © 2013 by the President and Fellows of The fi rst and most obvious lesson Harvard College. The contents refl ect the remain unresolved. Why did views of the author (who is responsible of this history is that America has spectacular booms and busts occur for the facts and accuracy of the research always been a nation of real estate herein) and do not represent the offi cial when and where they did? Were views or policies of the Taubman Center for speculators. Real estate speculation State and Local Government. buyers largely rational, or were their was an integral part of the “winning beliefs inconsistent with any sensible Taubman Center for State and Local of the west,” the construction of Government model of housing prices? What role 79 John F. Kennedy Street our cities, and the transformation of did credit markets play in fueling the Cambridge, MA 02138 American home life, from tenements 617-496-1722 boom or causing the bust? Finally, [email protected] to mini-mansions. A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

Figure 1: Mean Reversion in MSA Housing Prices 2000 - 2006 and 2006 - 2012

Source: Price data from the Federal Housing Finance Agency (FHFA) http://www.fhfa.gov/Default.aspx?Page=87

The second lesson is that these boom-bust prices even though low interest rates have cycles can generate signifi cant social costs, been less important than other factors in primarily through ensuing fi nancial chaos. This generating price booms. The Chicago land fact implies some urgency to rethinking the boom of 1836-37 was coincident with the national and local policies that impact housing chartering of two new state-supported Illinois markets. If buyers are particularly prone to banks. Securitization of mortgages for builders engage in wishful thinking about future price in the 1920s appears to have decreased the appreciation, then policies that encourage downsides of development, and increased credit homeowner borrowing can lead to larger social availability also boosted prices during the losses. recent boom. The third lesson is that the high prices paid The fi fth lesson is that the dominant mistake during the boom and the low prices paid made by investors is underestimating the during the bust are typically compatible with impact that elastic long-run supply of land, reasonable models of housing valuation and structures and crops will have on future land defensible beliefs about future price growth. values. Land buyers during Alabama’s 1819 Manhattan’s builders in 1929 could justify land boom look sensible given then-current their land purchases based on offi ce rents at the cotton prices and trends, but land values time. Real estate economists have examined depreciated as cotton prices fell with increased price-to-rent ratios in 2004 and argued that they U.S. and worldwide supply. Home buyers in seemed reasonable given plausible expectations Las Vegas and Phoenix in 2005 seem to have about future price growth and current capital misunderstood the almost perfectly elastic costs. supply of homes in America’s deserts. The fourth major lesson is that under-priced The sixth lesson is that the Great Housing default options can often help explain high Convulsion of the past 16 years is unlike 2 A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

previous booms in at least one major way. In boom. The fi rst is a “Gordonian” approach, every previous episode there was signifi cant which uses fi nance to determine the net present uncertainty about major economic trends that value of a property. The second is the “Von would impact land values and housing prices. Thunenite” approach, based on the Rosen- In the late 18th century, it was unclear how Roback economic model, which justifi es quickly transportation costs could fall and how prices by comparing local prices to the prices fast Americans would move west. The future in similar geographic areas, thus determining appeal of dense downtowns, like Chicago and if a given local price is “reasonable.” Buyers’ New York, was unsure in 1835 and still unsure primary error appears to be a failure to a century later. There is no obvious equivalent internalize Alfred Marshall’s dictum that source of uncertainty during the Convulsion. “the value of a thing tends in the long run

Limited Rationality and Housing Markets Surveys of home buyers during Before delving into more detail about the booms suggest that they hold historic housing booms, a framework of a dizzying array of apparently economic analysis is presented. Over the past inconsistent beliefs about future 20 years, spurred on by experimental evidence prices. Buyers in Boston in 2004, and events like the Internet equity boom and bust, economics has increased embraced for example, on average report “behavioral” models that assume either limited that they believe that housing cognition or downright irrationality. Surveys of prices will increase by 10.6 home buyers during booms suggest that they percent in each of the next ten hold a dizzying array of apparently inconsistent years, but only by 7.6 percent in beliefs about future prices. Buyers in Boston the next year. in 2004, for example, on average report that they believe that housing prices will increase by 10.6 percent in each of the next ten years, to correspond to its cost of production.” but only by 7.6 percent in the next year. These Frequently supply of land and housing is beliefs have many plausible sources. Some ignored, but land is abundant in the United buyers may be extrapolating from recent States, and even the most desirable cities can trends; others may be engaged in wishful be reproduced. Additionally, building up – thinking about the value of their largest asset. through skyscrapers – can provide a virtually “Entrepreneurs of error” may persuade buyers limitless supply of space. However, this error that home prices will experience dazzling is better seen as limited cognition— failing to future growth. The history of real estate use a sophisticated model of global supply and bubbles is replete with examples of interested demand – than as stupidity or irrationality. parties hyping local land values. Rural Speculation Though behavioral models are quite useful, economic models will lose all discipline if they 18th Century Land Speculation in New York treat investors as blank slates that irrationally State absorb any foolish notion that they hear. An Robert Morris occupied a storied place in alternative approach is to assume that such American history, as a merchant, fi nancier of beliefs are limited by sensible models of asset the revolution, signer of the Declaration of valuation. There are two reasonable models Independence, and national “Superintendent that can be used to justify prices during a of Finance” from 1781 to 1784, when he may 3 A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

have been the wealthiest man in America. in 1817 for $11,220 ($168,000 in 2012 dollars), Morris gambled in real estate throughout and 973,000 acres were sold in 1818 for $7.2 his career, but his truly immense real estate million ($130 million in 2012 dollars). A 270 began in 1790, when he spent percent increase in price during a single year is about $175,000 ($4.4 million in 2012 dollars) impressive. to buy about 1.3 million acres of New York In 1819, the boom busted, the country went into State land. and Alabama land values plummeted. Morris resold the land in 1791 for 75,000 Land buyers owed $21 million to the Federal pounds, or $343,800 (about $8.4 million in government in 1820, and $12 million of that 2012 dollars), making a substantial profi t. amount was due from Alabama itself. The He plowed his earnings into massive land government responded to these debts with purchases both in New York and in other various relief measures and it reduced the credit states, but faced increasingly diffi cult credit available for buying public land. conditions. Morris’ ability to meet his debts deteriorated, and he progressively mortgaged The Alabama boom and bust his property. Eventually, he was unable to meet illustrates a phenomenon that his obligations and become bankrupt. will reappear throughout these Morris’ land purchases were not at absurd real estate episodes: an under- “bubble” level prices, but rather at prices appreciation of the long-run that were quite low both relative to future power of elastic supply to push prices and relative to prices elsewhere in the prices downward. country. Credit market tightening helps explain Morris’ decline, but increases in easy credit do not seem to have fueled his earlier buying. These boom prices were not as unreasonable as However, there is a credit puzzle hidden in they might fi rst appear. Alabama land yielded the Morris story. The people who invested in large amounts of cotton, and prices of cotton Morris did not have the same upside potential and demand for cotton were high. When cotton that the equity owners did. Yet they lent prices fell, land prices followed. The Alabama at relatively standard interest rates, suggesting boom and bust illustrates a phenomenon that that, like recent purchasers of mortgage-backed will reappear throughout these real estate securities, they may have underestimated the episodes: an under-appreciation of the long- risks inherent in real estate speculation. run power of elastic supply to push prices The 1815-1819 Huntsville Cotton Land Boom downward. At current cotton prices, land prices in 1818 Alabama were justifi able. But Before the national fi nancial crisis known as since land was so freely available, in the U.S. the Panic of 1819, Huntsville, Alabama was and elsewhere, a smart investor might have the epicenter of the housing boom. It combined reasoned that prices would eventually fall so excellent cotton-growing soil with access to that land prices in Alabama would resemble the Tennessee River, which brought access land prices of similarly productive places to the Ohio River, the Mississippi River and, throughout the world. That logic would have ultimately, the Gulf of Mexico. Transportation made the land buyer of 1818 far warier about was the key to making frontier land valuable, paying so much for even prime Alabama land. and water was the key to transportation. 5,610 acres of public land in Madison County, The boom was not initiated by any change Alabama (which contains Huntsville) were sold in credit policies for public land, but instead 4 A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

fueled by optimism about uncertain economic subsequent price collapse led to fi nancial fundamentals, such as declining transport costs failures. and English demand for American cotton. Alabama’s land prices were not obviously The great Chicago boom and rational in 1818, but they weren’t obviously bust of the 1830s has been seen irrational either. Ex post, the Alabama as the epitome of a classic real speculators look foolish, but ex ante, there was estate bubble, as prices for land enough uncertainty to justify the buying; prices would have been reasonable as long as cotton on the edge of America rose from prices stayed high, and that was hardly such a essentially nothing to New York crazy thought. levels in six years. The 1900-1940 Iowa Wheat Land Cycle During the period from 1900-1930, prices Urban Speculation for rural land, particularly in Iowa, fi rst rose The Chicago Boom:1830-1841 dramatically, reaching historic heights in the early teens, before dropping during the The great Chicago boom and bust of the 1830s 1920s, almost ten years before the Great has been seen as the epitome of a classic real Depression. The price growth of land during estate bubble, as prices for land on the edge the fi rst decades of the 20th century was of America rose from essentially nothing to understandable, as both wheat yields and New York levels in six years. Homer Hoyt’s wheat prices were steadily increasing over One Hundred Years of Land Values in Chicago this period. In fact, national wheat prices had (1933) remains the indispensible resource for increased 34 percent (in real terms) from 1910- 19th century Chicago real estate. Focusing 1916. Unfortunately, wheat prices switched on his data estimates for land values in the from growth to decline in 1917, when they hit Chicago loop, prices per acre in 2012 dollars their 20th century peak. International supply were about $800 dollars in 1830, $327,000 per recovered after World War I and American acre in 1836 and $38,000 per acre in 1841. In production stayed high. Over the 1920s, the the aftermath of the bust, the Bank of Illinois growth in world wheat production appeared fi rst foreclosed on sizable real estate holdings to be seriously outpacing the growth in world and then declared bankruptcy in 1842. wheat demand, and prices of wheat, and The Chicago boom was vitally connected thus rural land values, plummeted. Though with the deep currents of America’s economic reasonable projections of increases in wheat development. The Erie Canal, which had prices, yields and lower transportation costs opened in 1825, gave Chicago access to could readily justify the high land values seen the East Coast, and the State of Illinois was during the boom years, those projections were digging the Illinois and Michigan Canal wrong, and farmers should have anticipated which promised to give Chicago access to the fall in prices that would eventually result the Mississippi River System. With these two from abundant supply. Still, it would be a canals, Chicago would sit at the epicenter of far-sighted farmer indeed who wouldn’t have America’s transportation network. Additionally, been optimistic given over a decade’s worth of comparsions with New York and Cincinatti positive price movements. However, across the suggest that prices were reasonable at the time U.S. as a whole, farm debt per acre increased of the boom. fi ve-fold between 1910 and 1920, and this 5 A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

Many authors—Hoyt among them— discuss dramatically, and the population of Los Angeles the contribution of the role of easy money after increased from six thousand to fi fty thousand 1835 to the boom and bust. Since the Bank of between 1885 and 1890. Migrants saw Illinois was a creature of state policy, and since benefi ts in the southern California climate, the the legislature pushed the bank to support real agricultural value of its land and the economic estate, it is certainly possible that the Bank was opportunity, created partially by real estate not charging appropriate interest rates given speculation. the probability of default. In 1837, there was Within Los Angeles, there was considerable widespread panic, and in May 29, 1837, the demand for rented residential and commercial Illinois banks suspended payments. As the space, and given the high rents charged by banks careened towards bankruptcy, Illinois’ landlords and the relatively cheap prices of land , like the canals which and construction prices seemed reasonable. were supposed to be fi nanced by the banks, Los Angeles prices could also be justifi ed stopped. using comparisons to other cities. Residential However, the optimists were vindicated in properties in Los Angeles cost 40 percent of the long run. Even the buyers of the most prices in Cleveland or Chicago, and the Times expensive tract in the loop in 1836, on repeatedly compared Los Angeles with San Dearborn Avenue near the Chicago River, Francisco and pronounced its own city cheap. experienced 3.6 percent real property value appreciation over the next twenty years. The Los Angeles boom was However,ex post justifi cation is dangerous. precipitated by the entry of the Chicago is studied precisely because it ended Sante Fe railroad into the Los up as a success. Yet numerous other never-built Angeles market, which caused communities that went through land boom the price of transport for people and bust cycles during the same period are not and goods to drop dramatically, well-understood. and the population of Los Angeles Los Angeles in the 1880s increased from six thousand to Los Angeles in the 1800s, the “Chicago of fi fty thousand between 1885 and the West,” experienced a substantial run-up 1890. in values during the 1880s and a subsequent reversal based on land value data from reported Prices declined after 1888, but Southern sales in the Los Angeles Times from 1882 to California continued to grow. Since aggressive 1889. The median price per square foot, in fi nancing was provided by sellers not banks, 2012 dollars, increases from 1.8 cents in 1882 there was no fi nancial crisis during the bust. to 2.8 cents in 1885. In 1886, the real price per Los Angeles did have a large boom-bust cycle square rises to 6.9 cents, and then 9.3 cents in and people who bought during the boom did 1887 and 18 cents in 1888, before the price lose money. Yet prices were also quite justifi ed returns to 12 cents in 1889. The 90th percentile given subsequent events, at least in the city price in 1888 is 70 cents per square foot. itself. The biggest losses were sustained by The Los Angeles boom was precipitated by investors in outlying boom-towns, who don’t the entry of the Sante Fe railroad into the seem to have focused on the virtually limitless Los Angeles market, which caused the price supply of space in greater Los Angeles, at least of transport for people and goods to drop relative to the demand during the 19th century. 6 A Nation of Gamblers: Real Estate Speculation and American History TAUBMAN CENTER POLICY BRIEFS

New York City: 1890-1933 in search of a six percent return, and those During the great boom of New York during investors may well have under-appreciated the 1920s, median price per square foot the value of the default option that they were increases from $2.70 per square foot in 1920 giving the building’s promoters. ( $31 in 2012 dollars) to $4 per square foot The one approach that would have managed in 1929, ($54 in 2012 dollars) based the data to predict the future more accurately is collected by Nicholas and Scherbina (2011). Marshallian. At 50 stories a building, there was Can these higher prices be reconciled with essentially an infi nite supply of upward space rational buyer beliefs? From an analysis of in New York and Chicago in the 1920s. After the costs and revenues of both a tenement 1929, prices plummeted during a great global purchaser intending to rent out rooms and a meltdown. Yet even if the had builder of skyscraper looking for sell offi ce not occurred, it is hard to see how peak 1920s space, both the standard tenement purchaser prices would have been sustainable. Before and the skyscraper builder of the late 1920s 1961, there were no effective height limits on could expect to receive a relatively good return building up, only setback requirements, and the on investment. Additionally, given the cost of amount of space that could have been added living and productivity of those in New York is considerable. Prices would have eventually city, the prices paid seemed reasonable when been squeezed down near construction costs, at compared with other cities of the time. least for skyscrapers, which would ultimately causes the price of land to also fall. The period between 1945 and 1960 would seem to be an ideal Metropolitan Speculation setting for a housing bubble. Post World War II: The Housing Bubble That The economy was resurgent Didn’t Happen after World War II and the Great The period between 1945 and 1960 would seem Depression. Household formation to be an ideal setting for a housing bubble. The soared during the baby boom. economy was resurgent after World War II and the Great Depression. Household formation Most strikingly, there was a soared during the baby boom. Most strikingly, revolution in mortgage fi nance, there was a revolution in mortgage fi nance, making it far easier to almost making it far easier to almost anyone to get a anyone to get a long-term, long-term, relatively low rate mortgage. relatively low-rate mortgage. Before the Great Depression, down-payment requirements averaged 50 percent, and bank loans had terms under fi ve years, with six As for the role of credit, it is more likely that percent interest rates. In the 1940s and 1950s, an under-priced default option played more of Federal programs, including the Federal a role in encouraging the speculative activities Housing Administration, the Veteran’s of builders. While some mega-buildings of the administration and the Federal National 1920s, including the Chrysler and Empire State Mortgage Association (Fannie Mae), enabled a Buildings were largely self-fi nanced, there was massive increase in credit availability. an impressive increase in the securitization business for property-backed securities. These Yet during the entire 1950-1970 prices securities were bought by ordinary investors, remained astonishingly fl at across America’s metropolitan areas. Though almost everywhere 7 TAUBMAN CENTER POLICY BRIEFS

experienced a signifi cant increase in prices, explain the shift. However, there was a major those prices were perfectly in line with shift in California’s housing markets in the the general increase in construction costs 1970s: new supply fell signifi cantly. In the in America during that time period. The early sixties, California was responsible for natural explanation for the missing boom in over a fi fth of the total number of permits in prices after World War II is that there was an the United States. But permitting dropped off enormous increase in housing supply over the signifi cantly after 1965, and the housing stock same time period. During the 1950s, America grew by only 32 percent in the 70s and just 21 permitted 11.84 million housing units, which is percent in the 1980s. roughly the same as America permitted during What caused this shift? Starting in the early the twenty-six years from 1920 to 1945. The 1960s, activists started using environmental post-World War II era demonstrated exactly arguments to justify barriers to new building. what textbook economics predicts should All major private developments became happen when robust demand meets relatively required to go through an environmental impact elastic supply. Quantities rose and prices stayed review process, and there were myriad local relatively fl at. regulations as well, such as 60 acre minimum lot sizes. Limits on supply would have driven In the early sixties, California was up prices in any case, but buyers seem to have responsible for over a fi fth of the been particularly optimistic about future price total number of permits in the growth, expecting increases in prices to top United States. But permitting 14% over the next decade. dropped off signifi cantly after The events after 1989 were typical for the 1965, and the housing stock grew ends of booms, as supply gradually increased by only 32 percent in the 70s and and prices gradually fell. Prices took a long time to reach bottom, but fi nally in 1996, real just 21 percent in the 1980s. prices in Los Angeles hit 62 percent of the peak level. The California boom and bust is the California in the 1970s and 1980s precursor to the great convulsion of the last 10 For the fi rst half of the post-war period, years. The earlier event featured real shocks California housing prices didn’t seem all that to housing supply and a somewhat limited different from prices elsewhere in the U.S. ability to provide abundant housing elastically, Between 1950 and 1970, housing values in the especially in a short time period. Across California metropolitan areas didn’t grow much metropolitan areas during this period, there was faster than in other American metropolitan a tight connection between inelastic housing areas. By contrast, between 1970 and 1990, supply and the extent of price appreciation. The price growth was signifi cantly higher annually prices during both the boom and the bust were in California than elsewhere. compatible with reasonable valuation models. The shift in California prices wasn’t rooted Those models just weren’t right. in changes in credit markets: real mortgage The Great Housing Convulsion between 1996 rates were rising over much of this time and 2012 period, and local economic conditions don’t The basic contours of the period from 1996 seem to have driven the price rise. Rising to 2012 are well known. Across the U.S. as a wages or productivity when compared to other whole, there was a 53 percent real increase in geographical areas are also not enough to housing prices between 1996 and 2006, which 8 TAUBMAN CENTER POLICY BRIEFS

was followed by a 28 percent decrease in real Easy credit is a common explanation for the values between 2006 and 2011. The boom was boom. In previous work, I have argued that not felt everywhere equally, as price growth credit market conditions cannot explain the occurred disproportionately in the warmest boom if buyers are rational. The changes in quarter of America’s metropolitan areas. interest rates were too small to justify such Moreover, there was enormous mean reversion price swings. Scholars also stress easier across areas, as shown in Figure 1. If a place approval rates and lower levels of down experienced 10 percent more price growth payment, but it is hard to assess the magnitude between 2001 and 2006, that place on average of these effects since it is impossible to control saw prices drop by nine percent relative to adequately for the changing characteristics of 2001 prices. mortgage applicants. While the price boom does not seem to be While the previous booms were explained by changing credit conditions, associated with dramatic episodes interest rates were low enough to justify prices of economic uncertainty, it is hard given the standard Gordonian model, especially to fi nd any comparable force in the given reasonable growth rates. Additionally, recent boom. The economy was buyer expectations were far more optimistic not growing particularly swiftly, than historic norms would justify. For example, nor was it obvious that there in 2005, the average Orange county buyer said that he expected 15.2 percent price increases in were any tectonic shifts in the each of the next ten years. Such beliefs seem geography of American enterprise. utterly implausible, but even if buyers expect fi ve percent perpetual growth, they would essentially be willing to pay an almost limitless While the previous booms were associated amount for a new house. with dramatic episodes of economic Just as a Gordonian approach could explain uncertainty, it is hard to fi nd any comparable the boom, a Thunenite approach can also force in the recent boom. The economy was help explain the Las Vegas phenomenon. It not growing particularly swiftly, nor was it seems plausible that some Las Vegas buyers obvious that there were any tectonic shifts in in 2003 noted that prices seemed extremely the geography of American enterprise. Some low, relative to California, and reasoned denser, older cities like New York and Boston that conditions weren’t all that different. were doing particularly well, but that can do This reasoning may explain their increased little to explain the boom in Las Vegas and willingness to pay. Additionally, a free, or inland California. The move to the Sunbelt under-priced, default option might also add was continuing during this time period, but considerably to the willingness to pay. During much of that appears to have been driven by this period, the mortgage insurance practices of unrestricted supply of new housing, which Federally-subsidized mortgage giants Freddie should not have boosted prices. Land buyers Mac and Fannie Mae provide the most natural may have thought that the supply of new land explanation for why borrowers might have surrounding Las Vegas was likely to contract, received an underpriced default option. but reasonable projections still suggest that there was more than enough desert space There were few obvious changes in economic for America to build enormous amounts of fundamentals that set off the bust. The economy housing. continued to grow strongly throughout most 9 TAUBMAN CENTER POLICY BRIEFS

of 2007, but the Case-Shiller index reached Booms end when these optimistic projections its peak in April 2006. Nor is it obvious that fail to materialize, at least in the short run, but credit markets conditions were tightening. in many cases, the shocks seem like they should Perhaps the most plausible explanation is that have been predictable to a forecaster with a slowing price growth led to a reassessment of Marshallian appreciation for the power of long- future price growth, which is often given as an run elastic supply. A suffi ciently well-informed explanation for the end of a speculative boom. buyer in Alabama in 1819 should have been able to expect that world-wide cotton supply Conclusion would push prices down, just like a skyscraper The housing convulsion that occurred between builder in 1920s Manhattan should have been 1996 and 2012 has many precedents in U.S. able to predict that abundant offi ce space history. Americans have been speculating should decrease rents dramatically. In the recent heavily on real estate for centuries, and vast boom, suffi ciently well-informed buyers in Las fortunes have regularly been won and lost. Vegas presumably should have recognized that Many things are similar between the most America’s incredible abundance of desert space would ultimately limit the long run value of To an economist with the benefi t homes on the urban fringe of that metropolis. of hindsight, the drop in cotton The diffi culties in forecasting the impact of prices after 1819 may seem highly supply are both understandable and hard to predictable, but why should that arbitrage. They are understandable, because have been true among cotton the cognitive requirements needed to forecast farmers on America’s frontier? the impact of global supply conditions on local property values are large. To an economist with the benefi t of hindsight, the drop in cotton recent boom and previous events. Rising prices after 1819 may seem highly predictable, prices are most strongly associated with but why should that have been true among optimistic expectations, and credit market cotton farmers on America’s frontier? conditions more typically played a supporting The ubiquitous nature of housing convulsions role. The optimistic expectations have been remind us that seemingly safe real estate justifi able based on recent experience and a investments can leave a gaping hole in bank simple capitalization formula (the Gordonian balance sheets when things go sour. The approach) and by Thunenite comparisons with tendency of markets to crash teaches that land prices or rents in other areas. under-priced default options can lead to large In the most recent boom, paying high prices social losses, especially because of fi nancial required an optimistic assessment of future meltdowns. This fact implies that there may price growth. Expecting a better future was be advantages if bank regulators recognize the also critical to the rural land boom on the New regular tendency of real estate values to mean York frontier in the 1790s, in Iowa in 1910, revert after booms. and in the urban booms of Chicago in the 1830s and Los Angeles in the 1880s and 1980s. In other cases, such as the Alabama land boom of 1819 and tenements in New York during the 1920s, prices were reasonable even if rents would stay constant.

10 TAUBMAN CENTER POLICY BRIEFS

Endnotes References 1This policy brief is based on the text of the Hoyt, Homer. 1933. One Hundred Years 2013 Ely Lecture presented at the American of Land Values in Chicago. Chicago: The Economics Association convention in San University of Chicago Press. Diego. The text of the lecture is available in Nicholas, Tom and Anna Scherbina. 2011. full in the forthcoming issue of the American “Real Estate Prices During the Roaring Economic Review. Twenties and the Great Depression.” UC Davis Graduate School of Management Research Paper No. 18-09.

11 TAUBMAN CENTER POLICY BRIEFS

RECENT POLICY BRIEFS

“A Nation of Gamblers: Real Estate Speculation “What Good Is Wealth Without Health? The and American History,” by Edward L. Glaeser Effect of Health on the Satisfaction Derived (Harvard University) May 2013. from Consumption,” by Amy Finklestein, Erzo F. P. Luttmer, and Matthew J. Notowidigdo (Harvard “First Degree Earns: The Impact of College Kennedy School) May 2009. Quality on College Completion Rates,” by Joshua Goodman (Harvard Kennedy School) September “Improving Health Care Quality and Values: 2012. Local Challenges and Local Opportunities,” by Katherine Baicker (Harvard School of Public “Healthy, Wealthy, and Wise: Improving the Health) and Amitabh Chandra (Harvard Kennedy Productivity of Massachusetts’ Health Care School) May 2009. Spending,” by Amitabh Chandra (Harvard Kennedy School) May 2012 “Urban Inequality,” by Edward Glaeser, Matt Resseger, and Kristina Tobio (Harvard University) “Which Places are Growing? Seven Notable March 2009. Trends from Newly Released Census Data,” by Edward Glaeser (Harvard University) March 2011. “Silver Bullet or Trojan Horse? The Effects of Inclusionary Zoning on Local Housing Markets “How Massachusetts Can Stop the Public-Sector in Greater Boston,” by Jenny Schuetz, Rachel Virus,” by Thomas A. Kochan (MIT Sloan School Meltzer, and Vicki Been (Furman Center for of Management) March 2011. Housing Studies) March 2009.

“Did Credit Market Policies Cause the Housing “The Greenness of Cities,” by Edward L. Glaeser Bubble?” by Edward Glaeser (Harvard University), (Harvard University) and Matthew Kahn (UCLA) Joshua Gottlieb (Harvard University), and Joseph March 2008. Gyourko (Wharton School) May 2010. “The Seven Big Errors of PerformanceStat ,” “Geography, Venture Capital, and Public by Robert D. Behn (Harvard Kennedy School) Policy,” by Josh Lerner (Harvard Business School) February 2008. March 2010. “The Rise of the Sunbelt,” by Edward L. Glaeser “What Makes a City Entrepreneurial?” by (Harvard University) and Kristina Tobio (Harvard Edward L. Glaeser (Harvard University) and Kennedy School) May 2007. William R. Kerr (Harvard Business School) February 2010.

Taubman Center for State and Local Government The Taubman Center and its affi liated institutes and programs are the focal point for activities at Harvard’s Kennedy School of Government that address urban policy, state and local governance, and intergovernmental relations. More information about the Center is available at www.hks. harvard.edu/taubmancenter. 12