In dit boek presenteert Kevin Erdmann een nieuwe kijk op de oorzaak van de financiële crisis van 2008. De crisis is volgens hem niet veroorzaakt door deregulering, speculatie, een goedkope geldpolitiek of een global savings glut, maar wel door allerlei beperkingen op het aanbod van woningen. David Henderson bespreekt hier dit belangrijk boek.
- Het tekort aan woningen uitgedrukt in cijfers:
As shown in figure 1-1, the number of new housing units per new resident was lower during the past 30 years than it had been for the previous 30 years. The stagnation in new construction is also confirmed by data on the housing stock. Figure 1-2 shows that the housing stock, adjusted for population,1 leveled off in the early 1990s and then declined. The slight rise in the number of units per adult during the boom after the turn of the century was not an aberration from past levels or growth rates. How could this be? And how did experts arrive at the conclusion that the problem with the economy at this time was an oversupply of homes?
- De scheeftrekking van het aanbod vond niet overal plaats:
Divide the 20 largest metropolitan areas into four general groups: Closed Access cities, Contagion cities, Open Access cities, and Uncategorized cities.
• Closed Access cities. The Closed Access cities are New York City, Los Angeles, Boston, San Francisco (with San Jose), and San Diego. These cities have a distinct signature— very low housing starts (even during expansions), high incomes, high rents (even relative to those high incomes), high rent inflation, high prices (even relative to rents), and large rates of out-migration of households with low incomes, especially during the housing boom.
• Contagion cities. The Contagion cities are Miami;3 Riverside, California; Phoenix; and Tampa, Florida. These cities experienced a brief period of sharply rising home prices, but they differ from Closed Access cities in the other respects. Their housing bubbles were caused by massive inmigration from the Closed Access cities. These cities had a refugee crisis of sorts. In this way, they are the mirror image of the Closed Access cities.
• Open Access cities. The Open Access cities are Dallas–Fort Worth,4 Houston, and Atlanta. These growing cities are able to build enough new homes to meet demand. They did not experience the massive migration inflows from Closed Access cities that overwhelmed the Contagion cities, so home prices remained moderate throughout the housing boom. Housing price trends in these cities tend to look much like housing price trends in most of the smaller regional metropolitan areas and rural areas.
• Uncategorized cities. The cities of Chicago; Philadelphia; Washington, DC; Detroit; Seattle; Minneapolis; St. Louis; and Baltimore don’t fit easily into the other groups. Some of these cities, like Washington, Seattle, and Chicago, are dealing with the same pressures that the Closed Access cities are, and housing costs there have increased. But at the metropolitan area level, housing starts in these cities tend to be much more generous than in the Closed Access cities, housing costs as a proportion of incomes are near national norms, and domestic migration out of the cities isn’t as extreme as for the Closed Access cities. Some of these cities have been dealing with struggling economies, so there isn’t as much pressure on housing supply. Yet before the recession, even Detroit and St. Louis, with their struggling economies, issued permits at a higher rate than the major Closed Access cities did. It should be pointed out that this housing crisis is an international phenomenon—especially in the Anglosphere—from Sydney to London to Vancouver, although it appears that the international cities where home prices have risen have generally expanded their housing stock more generously than the US Closed Access cities have. The international examples might be more comparable to Seattle than to our Closed Access cities.
- Bekeken op deze "gedesaggregeerde" manier speculatie en kredietexpansie kan niet de oorzaak zijn van de stijgende prijzen:
When the country was viewed as a whole, it appeared that irresponsible credit expansion had allowed new homebuyers to engage in wild speculation, causing home prices to veer wildly from long-standing norms relative to rents and to incomes. Viewed on an individual basis, however, few places in fact met this description. In some cities, rents were stable and so were prices. In others, rents were relentlessly rising and so were prices— and, it happens, so were local incomes. Credit markets have little to do with these differing housing markets. Limited access to opportunity does.
In other words, two things happened between the early 1990s and 2007, and we can see them on these graphs. First, in just a handful of cities, rents shot up far above the national average. (The dots for those cities moved sharply to the right.) Second, prices in those cities rose even more.12 The national median price/rent ratio was about 15×, so in a city where annual rent rose from $10,000 to $20,000, we might expect home prices to have risen from $150,000 to $300,000, in proportion. Instead, home prices in cities with rising rents rose at twice the rate that rents did. (The median home cost $15 for each dollar’s worth of rental value. Yet the slope of the regression line in the 2007 panel is 31×, which means that for each extra dollar of additional rental value, a homebuyer had to pay an additional $31.) A home in a city with median rent of $20,000 was now selling for about $450,000 instead of $300,000. Households now had to pay two premiums to gain access to these cities. The first premium was the higher rent. The second premium was the cost of the home, which now reflected both a higher present rental income and a premium for expected future rent inflation.13 Contrary to the cacophony of reports that home prices became unmoored from rents, the correlation between rent and price strengthened during the housing bubble. On a national scale, it appeared as if rents explained very little, leaving credit expansion as the obvious cause of higher home prices.14 But this impression is reversed when metro areas are viewed individually, and rent and income effectively explain home price behavior around the country. As rising rent in the Closed Access cities moved further from the national norm, it became a more important factor in the value of homes. The relationship shown here actually understates the rising rents and costs in the Closed Access cities, because the first response of households to rising rents is usually to compensate by reducing their real housing consumption.15 If households weren’t making this adjustment, the median rents and home prices of the Closed Access cities would have moved even further from the norm.16 Since the supply problem has only worsened since the collapse in mortgage credit markets, by 2018 rent had shifted even higher among the Closed Access cities, and it became an even more important factor in relative home prices among the metropolitan areas.
The United States didn’t have a bubble in the real valuations of homes. The Closed Access cities did. Most of the country, like Dallas and Atlanta, never left the long-term range of valuations. They were near the high end of the long-term range, which is reasonably explained by low real long-term interest rates. The rest of the rise in the national average real home price, above that level, was from the localized urban housing shortage in cities like San Francisco. The instability was local. It was a Closed Access problem.
- Dalende en lage rentevoeten kunnen niet de oorzaak zijn:
From the late 1970s to the mid-1990s, • real long-term interest rates rose, decreasing the intrinsic values of homes; and • the inflation premium in mortgage interest rates declined steeply, increasing the potential accessibility of credit. The net effect of these interest rate changes is shown in figure 1-12. Mortgage payments became much more affordable over this time period, in all cities. This should have allowed households to bid up prices. But instead, those lower rates led to lower mortgage payments because home prices declined relative to rents even while mortgage affordability was improving. It is real long-term interest rates that have the greatest impact on home prices. In other words, credit constraints due to high nominal interest rates don’t stop households from bidding home prices higher. Rather, they generally bid prices up to their intrinsic values based on the real interest rate.27 Moreover, intrinsic values are such a strong influence on price that prices were relatively strong compared to both rents and incomes in the late 1970s in spite of unusually unaffordable mortgage rates.
Then, from the late 1990s to the decade after 2000, nominal rates fell for two reasons: • Real long-term interest rates fell. The real rate on 30-year Treasury bonds fell from nearly 4% in 1998 to less than 2% by 2005, increasing the intrinsic values of homes. • Inflation declined slightly, increasing the potential accessibility of credit. In Open Access cities, mortgage affordability remained stable while home price/income and price/rent levels moved back up to levels similar to those of the late 1970s. This is the outcome we would expect if real long-term interest rates were the dominant influence. Prices reflect intrinsic value in Open Access cities. In these cities, households were not bidding home prices up to high levels because of affordable mortgage terms. Mortgage demand triggered new homebuilding, keeping prices low.
- De aanbod van woningen bepaalt de hoogte van de hypotheken niet de rentevoet:
The increased homebuilding boosted supply and pushed rents down, so price/rent levels increased partly because rents were lower. Equilibrium was reached as renting became more affordable. Both renters and owners in Open Access cities had more home at a lower expense. It would be difficult to overbuild in an Open Access city.28 Some places didn’t follow this pattern. In Closed Access cities where home prices were outrageous relative to incomes, households actually had to reduce their real housing consumption. All of their added costs stemmed from housing cost inflation, because of the bidding war for the limited number of existing homes. In Open Access cities, households were enjoying larger spaces. Open Access cities, where housing was abundant, were the sustainable cities. In Closed Access cities, undersupplied housing markets led to economic disruption. Note that, in the 1970s and 1980s, mortgage affordability followed a similar pattern in all cities. That’s because the primary factor in high mortgage payments was the impact of inflation on mortgage rates. But in the 1990s and in the decade after 2000, mortgage expenses were low in Open Access cities but high in Closed Access cities. That is because high mortgage payments were caused by lack of housing supply, not by interest rates.
- Oorzaak en gevolg werden verwisseld:
Causation can apply in both directions here. Loose terms from a deregulated credit market can lead to a supply of risky mortgages. And high prices in supply-constrained cities can lead to demand for risky mortgages. Clearly both of these factors were in effect to some extent during the decade after 2000. Deregulated credit markets have largely been blamed for destabilizing the economy, but it is causation in the other direction, where high costs lead to demand for risky borrowing, that is the more serious source of destabilization.
In fact, in the majority of the country the housing ATM story is a myth. Home prices were rising moderately. Homebuyers who found themselves unable to sustain their mortgage payments were defaulting in those cities. Even Phoenix and Las Vegas, Contagion cities, did not have particularly unusual home price behavior during the years preceding their sudden price bubbles in 2004 and 2005. In most of the country, government subsidies, loose money, generous banks, and even fraud among borrowers and lenders did not create price bubbles in housing. In the late 1970s, inflation made mortgage expenses high across the country. It was appropriate to address that problem with tighter monetary policy in 1980, and when that happened, mortgage expenses fell in all cities. In the decade after 2000, mortgage expenses were only high in Closed Access cities. The solution to that problem should have applied only to those cities. When policymakers attempted to solve the problem with contractionary national banking and monetary policies, mortgage expenses in Open Access cities, which were already at very low levels, shifted downward to even lower levels. In all cities, the cost of buying a home with a mortgage began to decline in 2006. This might have seemed like good news in the Closed Access cities, but this was bad news in the Open Access cities. This was the wrong solution.
The idea that credit expansion and private securitizations caused both rising homeownership and rising home prices is contradicted by the fact that the spike in private securitizations lagged homeownership even more than price increases did. As we can see in figure 3-5, when homeownership began to rise (point 1), 6% of residential mortgages were held by private pools, which include subprime and Alt-A loans. Market share had risen to more than 7% by 1998, when half the rise in homeownership had happened and home prices were still near long-term lows (point 2). Privately securitized market share was still only 9% of the total pool of mortgages by the end of 2003. When homeownership peaked in the second quarter of 2004 (point 4), the boom in private securitizations had just begun. By the end of 2006, just 18 months later, private securitizations had ballooned to 21% of outstanding mortgages. The boom in private securitizations came after the rise in homeownership. The explosion of private mortgage-backed securities and the later development of various versions of collateralized debt obligations (CDOs),16 which are widely cited as the cause of the price bubble, are clearly lagging indicators here. Strikingly, these securities are also unrelated to new homeownership. Even during the period where privately securitized loan rates were slightly elevated and homeownership was rising, little of the new ownership appears to have come from the net effects of subprime mortgage originations.17 The sharp jump in private securitizations and CDOs, if it is an indicator of anything, appears to be an indicator of the beginning of the bust.
One could say that the decline in first-time homebuyers after 2004 was a return to the pre-1995 stable trend, but this is a much different claim from saying that private securitizations created a surge of new homeownership that faded as the market was saturated. There was no massive influx of additional buyers when privately securitized mortgages jumped to 21% of the market. If credit expansion had been driving the housing market, we should have expected to see first-time homeownership rising, not dropping to the lowest level of the decade.
Broadly speaking, this rising price/rent ratio appears to explain why the value of low-priced houses in Closed Access cities rose relative to high-priced houses, and why that effect was much less common in cities that weren’t Closed Access. Low-priced neighborhoods appreciated rapidly in the Closed Access cities because they benefited from both rising rents and an increasing price/rent ratio. In contrast, the highest-priced homes in those cities had already reached the ceiling where price/rent levels off. It was the high-end home price/rent ratios rising more slowly than homes in the rest of the metro area that explains why lower-priced homes in those areas showed greater price appreciation on average than high-priced homes.
- Een huizenmarkt die op slot zit is ook een verklaring voor stagnerende lonen, stijgende ongelijkheid (volgens sommigen is de stijgende ongelijkheid overigens een oorzaak van de crisis), en stijgende winsten voor ondernemingen:
There has been a growing perception that wages are stagnant, incomes are becoming less equal, and national income has been accruing more to asset owners than to workers. This situation is frequently blamed on corporate power or market deregulation. But a significant cause of this economic malaise is the exclusion created by Closed Access housing. In fact, the Closed Access housing problem sheds light on how important free and open capital markets are for the well-being of the typical worker. Houses are essentially pure forms of capital— deferred consumption, the textbook definition of capital. The act of building a house is purely the act of forgoing the immediate consumption of goods and services to create shelter that will be consumed for many years into the future. The rental value of a home—the value that is consumed today—is almost purely a transaction between a consumer and capital, free of organizational intermediaries. Especially for homeowners, there is no network of firms and workers that must produce, administer, and deliver that shelter each month.
Furthermore, what the experience of Closed Access cities has made clear regarding this pure relationship between consumer and capital is that, first and foremost, where capital is obstructed from entering competitive markets, consumers are harmed. Especially consumers of limited means. The migration flows that have developed because of the Closed Access housing problem remove any doubt about this. Households with the lowest incomes systematically move away from cities where capital is obstructed from being used for housing and move to cities where capital flows relatively freely.1 This pattern is true of capital markets more generally, though it can be harder to see in most sectors, where the relationship between capital, laborer, and consumer is more complicated. Open Access economies tend to provide a larger share of income to laborers. Where rule of law is universal and freedom of contract is protected, capital has fewer systemic risks, and new entrants bid profits down. For instance, investors in developing markets tend to require higher expected returns on investment than investors in developed markets. Labor’s share of income in developed economies tends to be about 65%–70%. In developing economies, it tends to be about 55%–60%.2
The same pattern occurs within markets. Investors tend to demand lower returns for taking risks during long periods of expansion, such as the 1960s and 1990s. Unemployment also declines. Expected stability lowers the share of income earned by firms.3 When poor countries improve their local institutions to make the market for new capital more stable and safe, capital tends to flow to these economies, and wages rise. Because these economies begin with low wages, low wages are erroneously viewed as a competitive advantage that attracts capital. But capital doesn’t flow to where wages are low and will remain so. Capital flows to where wages are rising because local institutions are improving, productivity is rising, and ultimately the returns required by investors are declining. Where capital flows, wages rise and working people benefit. In this respect, Closed Access cities resemble less-developed economies.4 Capital income is bloated by limited entry. But because limited entry happens through the housing market in Closed Access cities, workers act as conduits for the bloated capital income—earning it first as gross wages before transferring it to landlords as rent. So migrants crossing America’s southern border from Latin America are attempting to increase their incomes, and migrants crossing the borders of America’s Closed Access cities as they leave are attempting to decrease their cost of living. But, fundamentally, the cause of the migration is the same. The lack of open and stable capital markets in the places these migrants are moving away from creates inequitable economic conditions.
The higher incomes in the Closed Access cities have three basic sources: 1. Composition. The economic segregation set into motion by Closed Access policies could mean that more-productive workers are more numerous in Closed Access cities. In other words, the median household in San Francisco may have more valuable job skills than the median household in Atlanta because Americans are self-segregating, according to income, into and out of Closed Access cities. 2. Productivity. There could be qualities about these cities that allow workers to be more productive. Networking potential for, say, tech designers in San Francisco or investment bankers in New York City is an example of how urban labor markets can increase productive opportunities for their workers. Economists sometimes call these “agglomeration economies.” 3. Monopoly. Limited entry can create monopoly profits. Here, the source of limited entry for workers is Closed Access housing. Changing composition could lead to higher incomes, but in that case, we would expect workers to continue to spend a normal proportion of their incomes on housing. Even in a city with some limits to housing expansion, households tend to adjust the location and amenities of their homes to keep housing expenses in a comfortable zone.6
When Rockefeller was consolidating the oil industry and Carnegie was consolidating steel, workers’ wages were rising a bit, but the primary effect of rising productivity was falling prices. In an open market, productivity flows largely to the consumer, even in cases where producers are powerful. So, while the composition of workers could lead to some rising incomes, rising local sources of productivity should mostly lead to consumer benefits. In a city without Closed Access limits to entry, new firms and workers would enter the market. Even if firms have monopolistic profits from patents or network effects, firms should retain those profits in a city where workers are free to move in and bid wages down.
The taxi industry is a good recent example of the power of limited entry versus productivity when it comes to wages. Uber and Lyft allow drivers to be much more productive. This has greatly increased the market for ridehailing services. But what the taxi companies had that Uber and Lyft didn’t were taxi medallions—government permission to operate in a field protected from competition. Those medallions used to be very valuable. Uber and Lyft bring two features to a local ride-hailing market—tools that make the service more productive and open access for potential new drivers. The combination of higher productivity and open competition lowers prices and increases quantity, but it does not increase wages.
This brings us to the third source of higher incomes: monopoly, or limited entry. Limited entry is the most powerful source of per sis tent excess income. It can ensure that the producers retain the new wealth that is created by higher productivity. Economists call this excess income “economic rents,” not to be confused with the rental value of a home. In the Closed Access cities, the productivity is created by the workers and the firms they work for, but it is real estate owners who earn excess income from limited access. In order for the real estate owners to collect their economic rents (which, here, do happen to come in the form of rental value on properties), the workers must first collect them as higher wages before transferring them to landlords as rent payments. As transfer agents, the workers are a cost to the firms and a benefit to the real estate owners. For limited access to be profitable, the real estate owners must have a source of geographically captured productivity that produces the income. So, fundamentally, it is a limit on the supply of residential space that creates the monopoly power, but that power is only profitable because it creates a limit on the supply of labor where that labor is productive enough and geographically captured enough to be able to pass on the higher costs to the firm and the consumer.
The Economic Policy Institute published an extensive report on the divergence between median pay and productivity growth since the 1970s.10 In this report, authors Josh Bivens and Lawrence Mishel analyzed a version of what Atif Mian and Amir Sufi have called “The Most Important Economic Chart” (figure 7-4).11 This chart shows the gap between productivity growth (the top line) and growth in median real hourly compensation (the bottom line) from 1973 to 2014. Working down from the top, Bivens and Mishel found that the loss in labor’s share of income accounted for only a small portion of the gap—only about 12% of the total (and this coming almost entirely in the period of dislocation after the 2008 recession). About 30% of the gap is explained by consumer inflation, which has been higher than producer inflation, which created a drag on workers’ real wages. And about 60% of the gap is due to a widening variance between the incomes of wage earners. Much of the divergence Bivens and Mishel found can be attributed to Closed Access housing policies. The plight of workers may be real, but it has almost nothing to do with a rising share of income going to employers. It has more to do with a rising share of income going to landlords in certain cities.
Bivens and Mishel found that very little of the gap between rising productivity and compensation of the median worker has been due to claims on that income by firms. This confirms the work of Matthew Rognlie of MIT and the Brookings Institution, who looked at capital’s share in many countries. He found that the rise in the share of income going to capital reflects gains in rental income earned by housing.12 As I have explained, this is true in the United States.
One reason that some have pointed to rising corporate incomes as a culprit for stagnant median incomes is that, recently, corporate profits have been high. But corporate profits are a poor indicator of total income to firms. As a share of national income, corporate profits can be high because corporate interest expenses are low. Interest expenses can be low either because interest rates are low or because firms are funding operations more with equity than with debt. Both of those factors are currently operative.
Over the long term, more firms have taken on the corporate form instead of being run as proprietorships. That also has increased the share of national income that is labeled “corporate profit.”13 Corporate profit is a poor indicator of the domestic earning power of firms.
A better way to measure the total operating income of firms over time is to simply consider the combination of proprietor income, after-tax corporate profit, and interest income. Figure 7-5 is a stacked line graph of these forms of income, going back to 1929, when Bureau of Economic Analysis data begin. Over the entire period, total capital income has been quite level—remarkably flat after adjusting for cyclical variations.14 In fact, it would take an economic upheaval larger than anything experienced in nearly a century for the split in capitallabor income to account for more than a few percentage points of difference in total labor compensation. It would be nearly impossible for a persistent shift in capital income to firms to be a significant cause of The Most Important Economic Chart.
Figure 7-6 compares shares of housing income, labor income, and other capital income to show the proportion of total domestic income going to each category. The most significant change since the 1970s has been a shift from labor income to housing income—most of which is taken by owner-occupiers and some of which is claimed by mortgage lenders and landlords.15 Total net capital income to homeowners and lenders has increased from about 4.0% to 7.4% of net domestic income over the 40 years since the 1970s.16 Much of that increase in income to housing has been a transfer to existing owners in Closed Access cities, through rent inflation on existing properties. This is not income that is increasing because more homes are being built. This is income that existing owners receive in cities where new homebuilders are locked out. This is income for not building. The only significant decline in the share of income to housing since the 1970s came when housing starts were strong. In the mid-1990s, as housing starts recovered from the contraction of the early 1990s, housing income leveled off at about 6.8% of net domestic income. From 2002 to 2006, at the height of the housing boom, when supply was pushing down rent inflation, housing income declined from 6.9% to 6.1%. After 2006, housing starts dropped sharply, and by 2009, housing’s share of domestic income was back to new highs.
Ironically, in order to reduce the income going to capital and increase the income going to labor, cities need to allow more capital investment in residential housing. More investment in Closed Access housing would reduce income to Closed Access housing. Every homeowner group fighting against a new apartment building and every monetary official warning that “hot” real estate markets could cause a destabilizing collapse in real estate prices understands this.
Their only error is in trying to stop it. Homebuilding is frequently cited as a way to raise compensation by creating construction jobs for the new units. But, more importantly, homebuilding increases real net compensation by reducing rents on all units and lowering the cost of living.
Furman and Orszag show how rates of return on investments have increasingly diverged among firms, with high returns generally being shared with labor. Workers who appear to have similar levels of skill and productivity earn higher incomes if they work at high-return firms. Plus, these high returns and wages within certain firms are persistent over time.20
The Closed Access housing problem imposes three significant costs on the domestic economy: 1. An indeterminable number of innovations simply haven’t happened because innovators are locked out of the creative mileu of the innovative skill centers. In 2017, 81% of all venture capital was invested in five cities— San Francisco; New York City; Boston; San Jose, California; and LA—the Closed Access cities.28 These are cities that, according to the Bureau of Economic Analysis, have general price levels about 19% higher than average, mostly attributable to housing costs.29 2. American consumers pay higher prices for goods and services that come out of these labor markets (finance, technology, biotech, etc.).30 The consumer surplus that should have been created by recent US innovation booms has been cut short because the wages of workers who have developed these innovations have been inflated in order to pass on excess returns to owners of urban real estate. 3. Economic opportunities are denied to struggling workers who would move to the Closed Access cities to work in service sectors. Thriving service sectors in these cities would attract spending from the high-income workers there who have significant disposable income. In a widely cited recent work, Chang-Tai Hsieh from the University of Chicago and Enrico Moretti from the University of California, Berkeley, estimate that housing constraints have reduced aggregate US GDP by 13.5%.31 Removal of these constraints in high-income cities, according to their model, would lead to an annual wage increase for the average worker of $8,775. More than two-thirds of this drag on domestic income is due to just two metro areas: New York City and the San Francisco– San Jose area. Highlight
The primary obstacle to labor mobility is the lack of housing where the new jobs are in the nontradable sectors. Ironically, the growth of construction spending in the Open Access parts of the country during the decade after 2000 is frequently cited as a problem—malinvestment. But, to the contrary, the primary problem was in the cities where real housing investment is severely limited. The malinvestment wasn’t due to building unneeded homes. It was due to not building homes where they were demanded.
In these passages, we can see that Autor, Dorn, and Hanson view the trade deficit as a factor delaying the transition of workers to new sectors, that neoclassical economic models fail to account for these delays in labor transition, and that the costs of these delays fall disproportionately on displaced workers. For the economy as a whole, the gains from trade are significant, but “the ultimate and sizable net gains are realized only once workers are able to reallocate across regions in order to move from declining to expanding industries.”51 This suggests that an overreliance on neoclassical economic models has harmed vulnerable workers and led to a backlash. But the key problem with the neoclassical economic models is that they assume that capital can be freely allocated to its best and most-valued use. The Closed Access cities have thrown a monkey wrench into a key adjustment mechanism of the economy. The solution to this isn’t to rid ourselves of neoclassical economic models. The solution is to rid ourselves of Closed Access housing obstruction.
- Vrijhandel heeft niet de gewenste voordelen opgeleverd precies omdat de huizenmarkt op slot zit:
Trade with China isn’t the fundamental problem here. It may have been one of the symptoms that exposed the problem. But the fundamental reason that economic growth and trade have not been broadly shared is the rise of Closed Access.
- GLobal savings glut als ontoereikende verklaring van de crisis:
First, American foreign direct investments earn higher returns than foreign investments in the US partly as a result of American organizational capital, expertise, and so on. Foreign investments into the US tend to be more weighted in low-risk debt securities. Second, the dollar provides stable liquidity in world markets, so US savers invest in productive foreign assets while some foreigners get liquidity value just from holding dollars. Third, because developing markets lack a stable source of safe investments, foreign savers are willing to invest in lowyielding, safe US assets while risk-seeking US savers buy riskier, high-yield foreign assets. In short, the US is selling knowledge, liquidity, and insurance in complex ways that don’t always show up on official balance sheets put out by the Bureau of Economic Analysis. These explanations are somewhat satisfying, and they go a long way toward explaining the mystery. Yet notice the coincidence that the scale of the gap surged at the same time that Closed Access took hold of the American economy. Beginning in the late ’90s, US foreign income shot up as our net foreign asset position dropped. American savers have long earned excess returns on foreign assets, compared to foreign investors in US assets.4 In other words, some of the sources Hausmann and Sturzenegger give for our net foreign income have been around for a long time. It could be that the explosion in international asset holdings—from about 50% of US GDP in the 1990s to about 150% of US GDP in the decade after 2000, for both liabilities and assets—simply expanded this advantage arithmetically. But I would suggest a fourth recent source for Hausmann and Sturzenegger’s dark matter: Closed Access rents and profits.5
Figure 8-3 is a slide from a presentation by Ben Bernanke. Bernanke (and Alan Greenspan) argued that a savings glut in the developing world was the causal factor bringing capital into the developed world, pushing down real interest rates, and thus pushing up property values. He is correct that this had little to do with monetary policy. But his explanation leaves many mysteries, including the first obvious question: Why did that capital flow to Australia but not to Germany? This is similar to the question that should be asked about the US housing bubble: Why did low interest rates or predatory lending lead to home prices doubling in coastal California but not in Texas? The answer to both of these questions is that the unstoppable force of postindustrial urbanization met the immovable object of constrained urban housing supplies in Australia and California, but not in Germany and Texas, and that was the development that set these global capital flows in motion.
- Speculatie en de ontwikkeling van nieuwe destabiliserende financiele instrumenten was een gevolg, niet de oorzaak:
The question is, what led to the rise of CDOs? They were created because there was a tremendous amount of capital in search of low-risk investments, and there weren’t enough willing or able mortgage borrowers to feed the frenzy of investors, so the existing low-rated MBSs were packaged up to create new pools that could issue new AAA-rated securities. The CDOs were used to create low-risk securities for investors because, without new mortgages, the MBSs couldn’t meet the demand. As the story goes, mortgages were being handed out to anyone with a pulse, and by 2007, in spite of lowering standards, the mortgage originators had run out of suckers. This would make perfect sense if the private securitization boom was actually built on the backs of borrowers who had little means to make their mortgage payments. But what if that story is wrong? What if the income of the typical homeowner had never declined, even when homeownership was rising? What if CDOs and most of the basic MBSs rose up amid a massive outflow of equity from the housing market?
In 2004, Karl Case and Robert Shiller, who are widely credited with calling the bubble and bust, had written, Judging from the historical rec ord, a nationwide drop in real housing prices is unlikely, and the drops in different cities are not likely to be synchronous: some will prob ably not occur for a number of years. Such a lack of synchrony would blunt the impact on the aggregate economy of the bursting of housing bubbles.13 If the bust they described had happened, the CDOs, which were vulnerable to nationwide generalized collapse, would have performed relatively well. When the collapse did spread across the nation, the reaction to these developments, because the under lying premise was wrong, was to insist on tightening access to mortgage credit, tightening monetary policy to reduce borrowing and investing, and accepting unprecedented declines in housing prices in order to impose discipline on the market. Yet the markets had shifted to a bias of fear and protection long before. That wasn’t widely appreciated because the CDO market was treated as if it was a product of risktaking— as if rabid demand for AAA-rated securities while short-term interest rates were hiked up and investment was collapsing was a sign of recklessness.
Descriptions of the CDO market are usually couched in language about leverage. But the global capital flows that were fueling this demand weren’t leveraged. Nor were the expanding domestic money market funds that were seeking safe assets. Investors were seeking secure cash flows. By the second half of 2006, mortgage growth was slowing. This is the period normally associated with declining lending standards. How can that be? If standards had become so lax, why were they attracting a declining number of borrowers? It is because this was the beginning of the crisis. Fear had already taken hold, and qualified borrowers were exiting home equity. It was the sharp decline in qualified borrowers that made the new pool of borrowers look worse, not an increase in less-qualified borrowers. And, because homeowners were cashing out and exiting the market well before the collapse of the CDO market, many of the homebuyers in 2006 and 2007 were investors rather than homeowners. Many of the early defaults were from investors, who are more likely to default tactically when prices fall than homeowners are. The reason so many of the recent buyers were investors is that homeowners had been selling their homes and either moving to less expensive cities or cashing out of homeownership altogether and sticking that cash into things like AAA-rated securities.