Apple bought Siri from its creator, SRI (Stanford Research Institute) International, which had developed it with government funding from the US Defense Advanced Research Projects Agency (DARPA) between 2003 and 2008. These earlier projects did not go viral; 2011 was the year in which, suddenly, people had a device in their pockets to talk with and to show off to almost-unbelieving admirers. Siri, and its soon-to-follow competitors, seemed to start the process of eliminating the need for human conversation. We might imagine preferring Siri as a conversation partner to a human, because Siri’s information is much more comprehensive and reliable. The idea that humans were ultimately replaceable was a scary thought, and it is easy to imagine a resulting loss of humanity’s collective self-esteem.
Around the same time, other inventions also attracted great public attention, notably driverless cars, which, despite some worries about safety, are predicted to replace many jobs. Though very few of us had actually seen a driverless car, we all knew that prototypes were already on our highways. These autonomous vehicles can already do things that we assumed were not programmable, like slowing down when the car senses children running around near the street. Human common sense can be reduced to a list of signals to a driverless car, which means that human common sense can be replaced.
Recent talk has stressed machine learning, in which computers are designed to learn for themselves rather than be programmed using human intelligence. A Google Trends search for Web searches for machine learning reveals a strong uptrend since 2012, with the Google search index more than quadrupling between 2004 and 2019. The narrative is propelled by recent stories. The highly successful chess computer program AlphaZero is described as working purely through machine learning—that is, without use of any human ideas about how to play chess. This narrative describes a tabula rasa program that plays vast numbers of chess games against itself, given no more information than the rules of the game, and learns from its mistakes.22 In some ways, the machine learning narrative is more troubling than computers running human-generated programs. Historian Yuval Noah Harari describes this narrative as leading toward a “growing fear of irrelevance” of ourselves and worries about falling into a “new useless class.”23 If they grow into a sizable epidemic, such existential fears certainly have the potential to affect economic confidence and thus the economy.
Of Jobs and Steve Jobs
The story of Steve Jobs is a remarkable narrative that ties into the fear of job loss to mechanization. His story was told in many books that appeared around the time of the 2007–9 world financial crisis. Particularly notable was the 2011 book Steve Jobs by Walter Isaacson, which sold 379,000 copies in its first week on sale,24 became a number-one New York Times best seller, and has over 6,500 reviews on Amazon with an average ranking of 4.5 stars out of 5. Isaacson specializes in biographies of geniuses (including Albert Einstein, Benjamin Franklin, and Elon Musk), but his book about Jobs was by far his most successful. Why did his book about Jobs go viral? Part of the answer was the timing: the publisher wisely dropped it into the market just weeks after Jobs’s death, allowing the news media narrative of his death to interact with the talk about the book.
Interestingly, the Steve Jobs narrative makes it appear that Jobs, a real human being with quirks that no one would program into a robot, was totally indispensable for Apple Computer. Jobs’s own story therefore became appealing to people who worry about their own possible obsolescence. He founded the company but was forced out, the story goes, because drab managerial types could not tolerate his eccentricities. When Apple began to fail, he was called back and breathed new life into the company, which is today one of the most successful in the world. The Steve Jobs narrative is a fantasy for people who don’t quite fit into conventional society, as many people with inflated egos but modest success in life may see themselves.
Economic Consequences of the Narratives about Labor-Saving and Intelligent Machines
We have traced much popular attention over two centuries to narratives about machines that will replace jobs. These narratives certainly affected, and continue to affect, people’s willingness to spend on consumption and investments, as well as their eagerness to engage in entrepreneurship and speculation. The economic hardships created by a temporary recession or depression are mistaken for the job-destroying effects of the machines, which creates pessimistic economic responses as self-fulfilling prophecies.
Henry George’s solution to the labor-saving machines problem—and the defining proposal of his book Progress and Poverty, published during the depression of the 1870s—was to impose a single tax on land, to tax away the labor-saving inventions’ benefits to landowners. George’s proposal assumed that the sole purpose of the new machines was to work the land, which might be the case if the economy is purely agricultural. This proposal is analogous to the much-discussed “robot tax” that appeared in public discussion during the Great Depression and has reappeared in the last few years. Taxing companies that use robots, the argument goes, will provide revenue to help the government deal with the unemployment consequences of robotics.25
George proposed to distribute part of the tax proceeds as a “public benefit.”26 His proposal is essentially the same universal basic income proposal that is talked about so often today:
In this all would share equally—the weak with the strong, young children and decrepit old men, the maimed, the halt, and the blind, as well as the vigorous.27
Other incarnations of the universal basic income proposal were offered by Lady Juliet Rhys-Williams in a 1943 book, Something to Look Forward To; a Suggestion for a New Social Contract, and by Robert Theobald in a 1963 book, Free Men and Free Markets. The Basic Income European Network (BIEN), an advocacy group, was founded in 1986 and later renamed the Basic Income Earth Network. The narrative that the future will be jobless for many or most people has helped sustain support for a progressive income tax and for an earned income tax credit, though in modern times it has not succeeded in producing a universal basic income in any country.
The mutating technology/unemployment narrative tends to attract public attention when a new story creates the impression that the problems generated by technological unemployment are reaching a crisis point. A celebrated 1932 book by Charles Whiting Baker, Pathways Back to Prosperity, sought to explain why the public’s concerns about labor-saving machines replacing jobs were wrong until now, the early 1930s. Baker emphasized the newness: “The widespread use of automatic machinery and economic transportation is only a thing of yesterday.” He stressed that unemployment was a new long-term problem, not going away, ever. Thus Baker advocated something like a universal basic income for all:
We have got to face the fact that there is one way, and only one, whereby we can make a market for our huge surplus of goods.… Increase the purchasing power of the 95 percent of the families of the United States who have only tiny incomes, and they will at once buy more.28
Recent years have seen a renewal of this great wave of concern as new redistribution proposals are put forth and discussed. Notably, Google Trends shows a huge uptrend in searches for the term universal basic income starting in 2012. ProQuest News & Newspapers reveals essentially the same uptrend. Public attention to inequality has burgeoned, with much attention to the increased share of income by the top 1% or the top one-tenth of 1%. Thomas Piketty’s Capital in the Twenty-First Century, which described this trend, was a best seller that generated intense discussion. The term “digital divide” has gone viral, describing a sort of inequality related to access to digital computers.
No one can predict the effects of labor-saving and intelligent machines on livelihoods and work in the future, but the narratives themselves have the potential to drive amplified economic booms and recessions, as well as public policy. The narratives at the time of this writing about artificial intelligence and machine learning replacing human intelligence and disintermediating skilled workers lend an instability to expenditure and entrepreneurship patterns. These and ot
her economic narratives may show up in the speculative markets, notably the real estate markets and the stock markets, to which we turn in the next two chapters.
Chapter 15
Real Estate Booms and Busts
Real estate narratives—stories about the often tantalizing increase in value of land, housing, locations, and homes—are among the most prominent economic narratives. A strong example of their influence was the talk leading up to the Great Recession of 2007–9, which disrupted economies all over the world. The 2007–9 Great Recession was fueled by stories communicating inflated ideas of the value of housing.
Real estate narratives have a long history. From ancient times through the Industrial Revolution, real estate talk centered on the price of farms. In modern times, attention shifted first to stories about empty city property suitable for building homes, then to actual homes in metropolitan areas. These shifts are just mutations of a perennial narrative about the scarcity of land and its value.
We might think that the real estate boom and bust narratives would be part of the same constellation of panic or confidence narratives that we discussed in chapter 10. But real estate confidence is very different from confidence in the state of the economy, because people tend to view the two as very different things.1 Real estate is regarded as a personal asset, which one might have useful opinions about, while the economy is seen as the product of myriad forces. As this chapter reveals, however, real estate is also a socially informed asset, with its value depending on how people compare themselves to their neighbors and beyond.
Speculation and Land Bubbles
For much of history before the twentieth century, popular narratives celebrated land speculation (either of farmland or of vacant city lots in burgeoning or promised cities) rather than home speculation or stock speculation. The following land speculator’s narrative, full of human interest, was written in 1840, after the collapse of a US land bubble that had started in 1837:
His father left him a fine farm free of incumbrance [sic]; but speculation became rife, fortunes were made in a twinkling, and D. fancied “one thing could be done as well as another.” So he sold his farm, and bought wild lands in the prairies, and corner lots in lithographed cities; and began to dream of wealth worthy of “golden Ind.” Work he could not: it had suddenly become degrading. Who could think of tilling or being contented with a hundred acres of land, when thousands of acres in the broad west were waiting for occupants or owners. D. was not the man to do it, and he operated to the extent of his means. At last the land bubble broke; lithographed cities were discovered to be mere bogs; and prairie farms, though the basis of exhaustless wealth, worthless unless rendered productive by labor.2
Here we see a perennial narrative of a foolish speculator buying unseen land in a bog, a narrative resurrected in the 1920s Florida land bubble, where a swamp replaced the bog.
The Florida Land Boom of the 1920s
There appears to have been little talk of single-family homes as speculative investments until the second half of the twentieth century. A ProQuest News & Newspapers search for home price reveals virtually no reference to the term in a speculative context until then. In fact, the phrase home price had a different meaning in past centuries, as in the home price of wheat, meaning the price of wheat in the domestic market as opposed to in foreign markets. When the phrase home price with its modern meaning was mentioned, it typically appeared in a story about a rich person spending a lot on a home, as a sign of wealth, but with no sense that the home was appreciating in value. For example, an 1889 article in the St. Louis Post-Dispatch exclaimed:
Senator Sawyer, who has for years lived in the house which Jefferson Davis occupied when he was here in Washington, has stopped paying rent and has built a MAGNIFICENT BROWN STONE MANSION within a stone’s throw of Dupont Circle. It is worth at least $80,000 and Sawyer’s millions will keep it in fine style. There are fine houses all around it.3
There is reference to value as if it is unchanging, but no sense that the senator might be making a speculative investment.
A ProQuest News & Newspapers search for price per acre shows a very different pattern. The phrase peaked at the beginning of the twentieth century, when it tended to refer to farmland as a speculative investment. The Florida land boom of the mid-1920s gets many hits, but the phrase home price almost never appears in those articles. During that widely discussed boom, an associated narrative emphasized that the proliferation of motorcars was making Florida land more easily accessible to northerners looking for winter homes. Given the rise of the automobile, it is not surprising that the allegedly beautiful sites that were selling out so fast were empty lots for building new homes. However, by 1926, the Florida land boom had become a widely covered scandal, reported nationally. Newspapers printed stories that promoters were selling undeveloped land divided into home-size parcels, sight unseen, to northerners who would never in their lifetimes see a town built near their isolated homes. These stories rendered such sales of undeveloped land disreputable.
Land has always been only a small part of a home’s value. One estimate, by Morris A. Davis and Jonathan Heathcote, suggests that the land’s value averaged only 36% of the home’s total value from 1976 to 2006.4 We do not seem to have data on the percentage of land value in home value for earlier years, except in assessments for property tax, but presumably when the US population was more rural, the percentage was even lower.5
In contrast to the Florida narrative, with its emphasis on land, investments in homes historically have been viewed as investments in structures that depreciate through weather and use, that require constant maintenance, and that go out of style and get torn down eventually. We can understand why land itself with no structure on it, at least during the Florida boom, seemed a more exciting investment.
Traditionally, prices of new homes were widely thought to be dominated by construction costs.6 In fact, it used to be conventional wisdom that home prices closely tracked construction costs. A 1956 National Bureau of Economic Research study noted some short-term movements in US home prices not explained by construction costs between 1890 and 1934, but it concluded:
With regard to long-term movements, however, the construction cost index conforms closely to the price index, corrected for depreciation.… For long-term analysis the margin of error involved in using the cost index as an approximation of a price index cannot be great.7
Because their construction cost index included only the prices of wages and materials, but not the price of land, the NBER analysts were viewing investments in homes as nothing more than holdings of depreciating structures, wearing out through time and tending to go out of fashion. With such a narrative, housing bubbles have little chance of getting started.
Enter News, Numbers, and Narratives
Newspapers eventually discovered that readers were interested in stories about home prices in congested inner cities, where the price of land is more connected with home prices because land is much more expensive there. These stories may have gained contagion, leading people to think that their properties far from city centers shared some of the same speculative trend to higher prices.
Another factor adding to contagion was the development of home price indexes for existing homes. The first mention of median prices of existing homes in ProQuest News & Newspapers appeared in 1957 in an Associated Press story referring to a US Senate housing subcommittee report, which concluded that low-income families were being priced out of the housing market partly because of the increased price of land.8 Newspapers began publishing the National Association of Realtors median price of existing homes in 1974. The Case-Shiller home price index (now the S&P/CoreLogic/Case-Shiller home price index), originally created by Karl Case and me, began to appear in 1991. These indexes allowed news media to regularly announce large movements, thereby lending concreteness to stories about movements in home prices.
Before the advent of statistical measures of home prices, it was relatively hard for the news me
dia to come up with regular stories about speculative movements in that market. Before stock price indexes became popular in the 1930s, writers for the news media were able to quote numbers illustrating big movements in the stock market, usually by quoting the one-day change in a few major stocks, which tended to move in the same direction on big move days. They lost no opportunity to write such stories. But it is not so easy to write about regular news in home prices. A house is almost never resold in just one day. Rather, most house sales occur over long intervals of time, years or even decades. Even changes in the median home price month to month were not newsworthy, because one-month changes could be erratic when different kinds of houses sold from one month to the next. The repeat-sales that Karl Case and I first started publishing in 1991 marked the beginning of a new era, one in which month-to-month changes in aggregate home prices could be inferred from highly disparate houses, each of which sells very infrequently. The indexes led to a futures market for single-family homes at the Chicago Mercantile Exchange that has the potential to reveal day-to-day changes in home prices, though activity on that market mostly dried up after the 2007–9 world financial crisis.
A common assumption in accounts of speculative bubbles in stock and housing markets has been that investors are extrapolating recently successful investment performance, expecting the price increases to continue and thereby eagerly forcing prices up even higher. This process repeats again and again in what may be called a vicious circle or feedback loop. However, narratives matter as well. If we listen to the narrative at such times, investors may seem a lot less calculating than they sometimes appear. Instead, the price increase appears to be driven less by future expectations than by the proliferation of stories and talk that draw attention to the asset that is booming, thereby fueling the bubble.
Narrative Economics Page 24