Farm employees in Washington State a difficult to find because many prefer UI to that work. (HT: MJPerry)
Thursday, May 27, 2010
Wednesday, May 26, 2010
Gender Segregation Around the Clock
Copyright, The New York Times Company
For many reasons, women are paid less than men. One is that many women have jobs with more desirable schedules.
Last week I wrote about how more women than men have been employed in payroll positions during the winter months of this recession, and how in April 2010 men regained their majority, with 50.04 percent of payroll positions. However, if you are reading this between 9 a.m. and 5 p.m., there are probably more women at work than men.
The chart below displays information about the work force by time of day, based on surveys conducted by the United States Census Bureau in 1997, 2001 and 2004 (when the 2010 Census is finished, an analysis like this will be possible for 2010). The calculations in the chart are based on work on gender segregation I have been doing with Christian Ferrada, a University of Chicago graduate student.
The black series shows the number of full-time people ages 18 to 65 who are at work in nonfarm jobs at each time of day, excluding people who work from home more than eight hours a week. Not surprisingly, it is vastly more common to be at work between 9 a.m. and 5 p.m. than to be at work in the evening or early morning.
The solid pink series displays the fraction of people at work at each time of day who were women, and the dashed pink shows the fraction for the entire day (weighting by the number of people at work). For example, only 32 percent of people at work at 7 a.m. were women, compared with 42 percent for the entire day.
An important reason for the time-of-day pattern of women’s share at work derives from the fact that men are more likely to have jobs that require early arrival at, or late departure from, work.
Interestingly, during the middle of the day the female share was 44 percent — two percentage points higher than the average for the entire day during the years studied. Given that 50 percent of payroll employees in April 2010 were female, we might expect that women in April 2010 would be as much as 52 percent of payroll employees at work at any time between 9 a.m. and 5 p.m. — a clear majority.
By the same logic, I expect that early in the morning or late in the evening women were nowhere a majority of employees at work, even during this recession when women were often a majority of payroll employees.
The vast majority of workers perceive work from 9 a.m. to 5 p.m. to be more desirable than work during the off-hours, and many of the off-hours workers are compensated with higher pay for the less desirable schedule. A variety of factors — including, some economists and many women’s rights advocates say, gender discrimination — may cause women to be paid less than men, but part of the reason may be the hours they choose to work.
For many reasons, women are paid less than men. One is that many women have jobs with more desirable schedules.
Last week I wrote about how more women than men have been employed in payroll positions during the winter months of this recession, and how in April 2010 men regained their majority, with 50.04 percent of payroll positions. However, if you are reading this between 9 a.m. and 5 p.m., there are probably more women at work than men.
The chart below displays information about the work force by time of day, based on surveys conducted by the United States Census Bureau in 1997, 2001 and 2004 (when the 2010 Census is finished, an analysis like this will be possible for 2010). The calculations in the chart are based on work on gender segregation I have been doing with Christian Ferrada, a University of Chicago graduate student.
The black series shows the number of full-time people ages 18 to 65 who are at work in nonfarm jobs at each time of day, excluding people who work from home more than eight hours a week. Not surprisingly, it is vastly more common to be at work between 9 a.m. and 5 p.m. than to be at work in the evening or early morning.
The solid pink series displays the fraction of people at work at each time of day who were women, and the dashed pink shows the fraction for the entire day (weighting by the number of people at work). For example, only 32 percent of people at work at 7 a.m. were women, compared with 42 percent for the entire day.
An important reason for the time-of-day pattern of women’s share at work derives from the fact that men are more likely to have jobs that require early arrival at, or late departure from, work.
Interestingly, during the middle of the day the female share was 44 percent — two percentage points higher than the average for the entire day during the years studied. Given that 50 percent of payroll employees in April 2010 were female, we might expect that women in April 2010 would be as much as 52 percent of payroll employees at work at any time between 9 a.m. and 5 p.m. — a clear majority.
By the same logic, I expect that early in the morning or late in the evening women were nowhere a majority of employees at work, even during this recession when women were often a majority of payroll employees.
The vast majority of workers perceive work from 9 a.m. to 5 p.m. to be more desirable than work during the off-hours, and many of the off-hours workers are compensated with higher pay for the less desirable schedule. A variety of factors — including, some economists and many women’s rights advocates say, gender discrimination — may cause women to be paid less than men, but part of the reason may be the hours they choose to work.
Monday, May 24, 2010
Wednesday, May 19, 2010
For the Moment, Men Regain the Majority
Copyright, The New York Times Company
By one measure, women were a majority of the nation’s work force for the five months ended March 2010. Employment trends suggest they will regain this majority next winter.
The chart below shows payroll employees by gender since January 2009. According to the Bureau of Labor Statistics, the majority of the nation’s payroll employees were female in February and March 2009. Women regained their majority status in November 2009 and held it until April 2009, when men held 50.04 percent of the nation’s payroll jobs.
Employment follows a seasonal pattern, and a somewhat different seasonal pattern for men than for women. Thus, the summer work force is more male than the winter work force, and it is no accident that women outnumbered men in colder months. So I expect that men will retain a narrow majority of payroll jobs until next winter.
The chart shows that the female majority was larger this winter than it was in the previous one, in large part because almost a million construction jobs were lost in 2009, and the construction industry employs many more men than women. The construction industry will not shed so many jobs this year, but it will take more than a year to recreate the construction and other male-dominated jobs that were lost in the previous year.
For this reason, I expect that women will again hold a majority of payroll jobs for several months next winter.
By one measure, women were a majority of the nation’s work force for the five months ended March 2010. Employment trends suggest they will regain this majority next winter.
The chart below shows payroll employees by gender since January 2009. According to the Bureau of Labor Statistics, the majority of the nation’s payroll employees were female in February and March 2009. Women regained their majority status in November 2009 and held it until April 2009, when men held 50.04 percent of the nation’s payroll jobs.
Employment follows a seasonal pattern, and a somewhat different seasonal pattern for men than for women. Thus, the summer work force is more male than the winter work force, and it is no accident that women outnumbered men in colder months. So I expect that men will retain a narrow majority of payroll jobs until next winter.
The chart shows that the female majority was larger this winter than it was in the previous one, in large part because almost a million construction jobs were lost in 2009, and the construction industry employs many more men than women. The construction industry will not shed so many jobs this year, but it will take more than a year to recreate the construction and other male-dominated jobs that were lost in the previous year.
For this reason, I expect that women will again hold a majority of payroll jobs for several months next winter.
Wednesday, May 12, 2010
Was There Good Reason for a Housing Boom?
Copyright, The New York Times Company
Several factors regarding housing supply and demand have changed since the turn of the century and will change in the years to come, in the direction of increasing the nation’s housing inventories. As a result, at least part of the construction boom of the last 10 years served a legitimate economic purpose.
Motivated in large part by extraordinarily high housing prices, the nation’s home builders were busy building and renovating from 2000 to 2006. As a result, the housing inventory at the end of the housing boom was 3 or 4 percent above the trends of the previous decade.
One perspective says that most of the housing boom was a colossal market mistake: an unfortunate consequence of easy money, lax regulator oversight, subsidies to home ownership and irrationally exuberant buyers. In this view, public policy would ideally have pushed back, and housing construction would not have boomed anywhere close to the way it did. Let’s call this the “bubble theory.”
The opposing view says that there were good supply and demand reasons to significantly increase the amount of housing above previous trends. With the advantage of hindsight, we might have preferred to have stretched out the 2000-2006 construction activity over a longer period, or to have put some of the new houses in different locations. Still much of the housing built in the 2000s will ultimately prove desirable. Let’s call this the “fundamentals theory.”
The various supply and demand forces can be distinguished according to whether they affect the amount of rent paid by a home’s occupants, or whether they affect the fraction of the rent that a landlord and his creditors can keep as a return on their capital investment. (For the sake of exposition, my discussion treats the landlord and tenant as distinct, even though with owner-occupied dwellings they are the same.)
The pie chart below shows the distribution of tenants’ rental payments to various uses.
The largest piece of the pie – “cash flow to owners and creditors” – is the amount going to the owner of the home or apartment, and his creditors, that serves as a return on investment. It includes the principal and interest paid on mortgages, and any surplus the owners may have after making a mortgage payment.
The other pieces include property taxes, payments to employees who manage and repair the property, and payments for intermediate inputs such as real estate agents and mortgage industry employees. (This last piece does not include principal and interest paid on mortgages, but would include the mortgage closing costs that finance the salaries of industry employees).
The “cash flow” piece is an important determinant of housing prices, because the “fundamentals theory” says that homes sell for the present value of the cash flow. The cash flow to owners, and therefore housing prices, would increase if rents increased, so that the total pie expanded. But the owners would also get more, and homes would be worth more, if owners could pay less to real estate agents, property managers, and mortgage brokers.
In other words, housing prices can also go up because the entire pie is expected to increase in size, or because the cash flow piece is expected to get larger while some of the other pieces get smaller.
An extraordinary population growth rate would be good reason to build houses at a faster pace and would cause prices of housing to increase for occupants. Yet the human population of the United States grew at about the same rate 1995-2005 that it did 1980-95 (about 1 percent per year during both periods, although the total pounds of humans have probably grown more in the recent period).
But let’s not forget about the non-humans in our homes – the stuff!
Consumer durable statistics from the Bureau of Economic Analysis confirm the obvious: From 1995 to 2005, ownership of furnishings and appliances increased at a rate of 4.5 percent per person per year (as compared to 2.2 percent in the earlier period), and ownership of recreational goods rose 9.1 percent per person per year (as compared with 5.4 percent earlier).
In case you still doubt that Americans have genuinely increased their demand for space, note that the self-storage industry has more than doubled its share of the national economy since 1995, according to employment data from the Bureau of Labor Statistics.
Another way to distinguish supply and demand forces is whether they have their effects on owner net cash flows before 2005, or were expected to affect those flows after 2005. To the extent that houses were more valuable during the boom because of favorable cash flows – perhaps because home occupants had a high demand for space to keep their stuff – rents probably would have risen faster than home prices, because home prices are a kind of average of current and future cash flows.
According to the Bureau of Labor Statistics, inflation-adjusted housing rents did increase in the late 1990s and early 2000s, but much less than home prices did. This is why I, and several other economists, have concluded that the majority of the housing boom derived from market expectations about the future, rather than something that happened to housing demand before 2006 (this same logic says that home prices were only moderately stimulated by the fact that mortgage rates were lower in the 2000s than they were in the 1990s: see these items from Professor Paul Krugman and me for more on that point).
Undoubtedly, Americans will continue to accumulate stuff, and will have commensurate increases in their housing demand, which could partly explain why housing prices increased more than housing rents during the boom. But the accumulation of stuff by itself would be unlikely to produce the housing bust we saw after 2006 because our furniture, clothing, toys, etc., did not suddenly disappear in 2007, but housing prices came crashing down.
That brings me to a second set of housing market factors that reinforce the increased demand for space: anticipated advances in information technology. Information technology will likely reduce home ownership costs – shrink the “real estate, mortgage industry employees, and other intermediate inputs” and “compensation of employees” pieces of the pie – but for now we cannot be sure exactly when and how much.
House shopping began an ongoing transformation over the past several years, with shoppers able to take spectacular virtual tours of properties. Arguably information technology will someday enable most home shoppers to do without a human real estate agent, and this would be a significant cost savings for home sellers. Don’t forget that a home is built once but typically sold many times, with each sale creating a fee for real estate agents, and many of the sales involving a vacancy period during which no one gets value from the structure.
Market participants might have also hoped that mortgage loan screening could improve, and become less labor intensive, thanks to information technology. The economists Robert Hall and Susan Woodward pointed out in 2008 that “Recent years have seen great improvements in data, especially the introduction of credit scores, which gave lenders new powers to forecast mortgage defaults and to adjust interest rates offered to prospective borrowers. In 1990, credit scores were rare; by 1996, they were standard.”
All of these things are expected to someday significantly reduce the costs of owning a home, managing it and keeping it occupied. But it will be years until we know exactly when and how much; in the meantime, the housing market may gyrate as it adapts to these historically rare, but ultimately beneficial, changes.
Housing bubble-theorists have not yet seriously considered factors like these. Until they have, it is too early to blame a majority of the housing boom on irrationally exuberant home buyers, because even without these things a historically unusual housing boom may well have been efficient.
Several factors regarding housing supply and demand have changed since the turn of the century and will change in the years to come, in the direction of increasing the nation’s housing inventories. As a result, at least part of the construction boom of the last 10 years served a legitimate economic purpose.
Motivated in large part by extraordinarily high housing prices, the nation’s home builders were busy building and renovating from 2000 to 2006. As a result, the housing inventory at the end of the housing boom was 3 or 4 percent above the trends of the previous decade.
One perspective says that most of the housing boom was a colossal market mistake: an unfortunate consequence of easy money, lax regulator oversight, subsidies to home ownership and irrationally exuberant buyers. In this view, public policy would ideally have pushed back, and housing construction would not have boomed anywhere close to the way it did. Let’s call this the “bubble theory.”
The opposing view says that there were good supply and demand reasons to significantly increase the amount of housing above previous trends. With the advantage of hindsight, we might have preferred to have stretched out the 2000-2006 construction activity over a longer period, or to have put some of the new houses in different locations. Still much of the housing built in the 2000s will ultimately prove desirable. Let’s call this the “fundamentals theory.”
The various supply and demand forces can be distinguished according to whether they affect the amount of rent paid by a home’s occupants, or whether they affect the fraction of the rent that a landlord and his creditors can keep as a return on their capital investment. (For the sake of exposition, my discussion treats the landlord and tenant as distinct, even though with owner-occupied dwellings they are the same.)
The pie chart below shows the distribution of tenants’ rental payments to various uses.
The largest piece of the pie – “cash flow to owners and creditors” – is the amount going to the owner of the home or apartment, and his creditors, that serves as a return on investment. It includes the principal and interest paid on mortgages, and any surplus the owners may have after making a mortgage payment.
The other pieces include property taxes, payments to employees who manage and repair the property, and payments for intermediate inputs such as real estate agents and mortgage industry employees. (This last piece does not include principal and interest paid on mortgages, but would include the mortgage closing costs that finance the salaries of industry employees).
The “cash flow” piece is an important determinant of housing prices, because the “fundamentals theory” says that homes sell for the present value of the cash flow. The cash flow to owners, and therefore housing prices, would increase if rents increased, so that the total pie expanded. But the owners would also get more, and homes would be worth more, if owners could pay less to real estate agents, property managers, and mortgage brokers.
In other words, housing prices can also go up because the entire pie is expected to increase in size, or because the cash flow piece is expected to get larger while some of the other pieces get smaller.
An extraordinary population growth rate would be good reason to build houses at a faster pace and would cause prices of housing to increase for occupants. Yet the human population of the United States grew at about the same rate 1995-2005 that it did 1980-95 (about 1 percent per year during both periods, although the total pounds of humans have probably grown more in the recent period).
But let’s not forget about the non-humans in our homes – the stuff!
Consumer durable statistics from the Bureau of Economic Analysis confirm the obvious: From 1995 to 2005, ownership of furnishings and appliances increased at a rate of 4.5 percent per person per year (as compared to 2.2 percent in the earlier period), and ownership of recreational goods rose 9.1 percent per person per year (as compared with 5.4 percent earlier).
In case you still doubt that Americans have genuinely increased their demand for space, note that the self-storage industry has more than doubled its share of the national economy since 1995, according to employment data from the Bureau of Labor Statistics.
Another way to distinguish supply and demand forces is whether they have their effects on owner net cash flows before 2005, or were expected to affect those flows after 2005. To the extent that houses were more valuable during the boom because of favorable cash flows – perhaps because home occupants had a high demand for space to keep their stuff – rents probably would have risen faster than home prices, because home prices are a kind of average of current and future cash flows.
According to the Bureau of Labor Statistics, inflation-adjusted housing rents did increase in the late 1990s and early 2000s, but much less than home prices did. This is why I, and several other economists, have concluded that the majority of the housing boom derived from market expectations about the future, rather than something that happened to housing demand before 2006 (this same logic says that home prices were only moderately stimulated by the fact that mortgage rates were lower in the 2000s than they were in the 1990s: see these items from Professor Paul Krugman and me for more on that point).
Undoubtedly, Americans will continue to accumulate stuff, and will have commensurate increases in their housing demand, which could partly explain why housing prices increased more than housing rents during the boom. But the accumulation of stuff by itself would be unlikely to produce the housing bust we saw after 2006 because our furniture, clothing, toys, etc., did not suddenly disappear in 2007, but housing prices came crashing down.
That brings me to a second set of housing market factors that reinforce the increased demand for space: anticipated advances in information technology. Information technology will likely reduce home ownership costs – shrink the “real estate, mortgage industry employees, and other intermediate inputs” and “compensation of employees” pieces of the pie – but for now we cannot be sure exactly when and how much.
House shopping began an ongoing transformation over the past several years, with shoppers able to take spectacular virtual tours of properties. Arguably information technology will someday enable most home shoppers to do without a human real estate agent, and this would be a significant cost savings for home sellers. Don’t forget that a home is built once but typically sold many times, with each sale creating a fee for real estate agents, and many of the sales involving a vacancy period during which no one gets value from the structure.
Market participants might have also hoped that mortgage loan screening could improve, and become less labor intensive, thanks to information technology. The economists Robert Hall and Susan Woodward pointed out in 2008 that “Recent years have seen great improvements in data, especially the introduction of credit scores, which gave lenders new powers to forecast mortgage defaults and to adjust interest rates offered to prospective borrowers. In 1990, credit scores were rare; by 1996, they were standard.”
All of these things are expected to someday significantly reduce the costs of owning a home, managing it and keeping it occupied. But it will be years until we know exactly when and how much; in the meantime, the housing market may gyrate as it adapts to these historically rare, but ultimately beneficial, changes.
Housing bubble-theorists have not yet seriously considered factors like these. Until they have, it is too early to blame a majority of the housing boom on irrationally exuberant home buyers, because even without these things a historically unusual housing boom may well have been efficient.
Tuesday, May 11, 2010
The Rise of the Self-Storage Industry
Housing bubble theorists will offer you a list of reasons why it might have been appropriate to have a bit of a housing boom, and then (perhaps correctly) show you that the factors on their list are not big enough to explain all of the housing boom we actually experienced. The final step in their argument is to attribute all of the unexplained part of the housing boom to irrational exuberance.
Of course, the entire argument falls apart if their list is incomplete. For example, do those lists include this?
American's incredible accumulation of stuff is real (itself caused by China's seizing its comparative advantage in making that stuff), and the accumulation undoubtedly increased American's demand for space at home. The rise self-storage industry is a reflection of that increased demand for space.
Note that I do not claim that the entire housing boom is due to American's demand for closet space. In fact, that's the point: there are any number of things that we may someday understand to explain the residual, and I see no reason why those possibilities are limited to "irrational exuburance."
Of course, the entire argument falls apart if their list is incomplete. For example, do those lists include this?
American's incredible accumulation of stuff is real (itself caused by China's seizing its comparative advantage in making that stuff), and the accumulation undoubtedly increased American's demand for space at home. The rise self-storage industry is a reflection of that increased demand for space.
Note that I do not claim that the entire housing boom is due to American's demand for closet space. In fact, that's the point: there are any number of things that we may someday understand to explain the residual, and I see no reason why those possibilities are limited to "irrational exuburance."
Monday, May 10, 2010
Unemployment Insurance and Unemployment
I have written before about statistical analysis of the relationship between Unemployment Insurance and unemployment. Here are some anecdotes along the same lines.
Sunday, May 9, 2010
Friday, May 7, 2010
This Week's Labor and Consumption Data Compared with My Forecasts
I made these forecasts last fall, and this week we had the release of another month of consumption data (Mar 2010) plus another month of labor data (Apr 2010). (Recall that the "partial recovery" scenario -- green series in Figure 5b -- is the one I thought would actually happen, in part because it was the only one consistent with consumption patterns in 2008).
Men Outnumber Women on Payrolls for the First Time in Six Months
Men are now 50.04 percent of payrolls, up from 49.82 percent in March.
Thursday, May 6, 2010
The Housing Cycle and Prospects for Technical Change
Information technology has dramatically advanced over the years, and it is unreasonable to assume that today it will suddenly cease. But what types of production will be most transformed by future advances, when will the gains will be realized, and in what magnitude? This paper suggests that housing is one sector to be significantly impacted by technical progress, and that some part of the housing boom and bust 2000-2006 was a rational and efficient anticipation of such events.
The prospects for technical progress are exciting, but they are inherently uncertain. We do not yet know how to cure cancer, or how to make a cheap personal computer run at 100 Ghz. Nevertheless, we would be ill-advised to make investment decisions today under the assumption that lifetimes will remain constant, or that personal computers will always run at the same speed. Some of the inputs that are complementary with future technologies – those that are inelastically supplied in the short run – should begin to be accumulated today, even though the progress on which they rely has not yet occurred.
If the new technologies are realized earlier – or in greater magnitude – than originally anticipated, then the complementary inputs will enjoy a favorable and abnormal return. If the new technologies are realized later, or in lesser magnitude, then the complementary inputs will be under-utilized for a while and thereby earn an unfavorable return. This paper modifies the existing economic theory of housing investment to include these ideas. I conclude that at least part of the housing boom occurred because technology was expected to advance in ways that would make existing housing structures more productive.
When a bust occurs as a consequence of bringing the future into better (and less exuberant) focus, the boom time construction was, in hindsight, too much and too soon, but nonetheless the housing stock might not return to its pre-boom trend. As long as there was at least partial truth in the original anticipations – i.e., at least some productivity advance really did occur – then part of the construction boom was necessary and efficient, even with the benefit of hindsight. Housing prices must crash during the bust, but net housing investment may continue to be positive during the bust period. A “bubble” – that is, an increase in housing prices that occurred for no fundamental reason – would not have these properties. Thus, a bust period that keeps the housing stock and housing prices above their pre-boom trends means that housing fundamentals are in fact better, and thereby rejects the pure bubble theory.
Figure 1 shows the basic price and investment patterns of housing boom and bust to be explained. The left scale shows two monthly measures of the average value of housing properties, known as the “Case-Shiller Home Price Index” and the “OFHEO Home Price Index” divided by the PCE deflator and normalized to 100 in the first month shown (January 2000 [note seasonal adjustment]). The right scale shows residential construction spending, expressed as a percentage of personal income. The bottom series measures tenant-occupied housing construction and the top series adds owner-occupied construction. The Figure shows that real residential construction grew as housing values did through 2005. Sometime between late 2005 and early 2006, both values and construction spending peaked and then declined through early or mid 2009. Since then, construction spending increased a bit, as did one of the measures of real home values (both value measures agree that the 2009 change was significantly less than the changes during 2007 and 2008). The Figure also shows that the large majority of the amount and changes in residential construction spending are for owner-occupied structures.
Figure 2 displays quantity indices (2000 = 100) for investment in three types of structures: owner-occupied residential, tenant-occupied residential, and non-residential. The three types of structures investment stayed in about the same proportions 1993-2000. Both types of residential investment increased significantly 2001-2005, and thereafter fell back to pre-boom amounts. In contrast, investment in non-residential structures fell after 2000, and increased significantly 2005-2008. Because the boom increased both types of residential investment and reduced non-residential structures investment, the question addressed in this paper is why housing prices and construction increased 2000-2005, without particular attention to the distinction between owner-occupied and tenant-occupied housing.
The time patterns for housing prices, investment (both residential and non-residential), and income flows qualitatively match my stochastic technology model, and are unlike market responses to a short run shift in housing demand. Thus, the bulk of the housing boom appears to be a response to expectations about the future, and the housing bust a revision of those expectations. Some of those expectations are related to technical change, although it is also possible that additional expectations related to the anticipation of government bailouts, and even irrational exuberance, added to – or even multiplied – the effects of prospects for technical progress.
Section I explains how much of housing output goes for banking services, real estate brokerage, and management, as opposed to the owner of the structure. Structures have a leveraged claim on housing output because they are less elastically supplied than are the housing sector’s other inputs. Moreover, prospects for technical progress in the provision of the elastically supplied inputs should increase both housing prices and housing construction, especially for types of houses that are relatively intensive in the elastically supplied inputs. Section II amends the familiar “q-theory” model of housing investment with an elastically supplied intermediate goods sector. I measure time series for the net operating surplus and vacancy rates emphasized by the theory, and show how they suggest that optimism about the future fueled the housing boom, rather than tastes, technologies, or subsidies during the boom. Section III therefore looks at the market dynamics that would be an efficient response to good news about the future. It shows the expected size of the housing service demand shift – that is, the amount that the housing stock is expected to increase in the long run – that is implied by a given amount of short run housing price elevation. It shows the quantitative expectations about tastes and technology that are required to generate large housing price increases in the short run and housing stock increases in the long run, and how to distinguish the theory from alternative theories of the housing cycle, leaving future research to answer the question of whether such expectations were reasonable. Section IV points out that real housing prices are still higher than they were before the housing boom, despite the boom’s apparent legacy of overbuilding, which suggests that some of the boom’s optimism still remains three years later.
The housing boom that lasted through 2006 featured high and rising real housing prices, high rates of real investment in both owner-occupied and tenant-occupied housing, and low rates of real investment in non-residential structures. Housing sector net operating surplus – that minority part of housing service rents that accrues to homeowners and their creditors – was low relative to the housing stock during the boom, as were occupancy rates. More recently, real housing prices are much lower, although probably still above their pre-boom levels. Housing investment is now below, and non-residential structures investment above, pre-boom levels.
It is sometimes suggested that the housing boom derived largely from low interest rates, and public policies promoting homeownership, in the mid 2000s. But low interest rates by themselves should push up all kinds of structures investment, not just residential structures. And policies to encourage homeownership should have depressed construction of rental housing units, when in fact that construction boomed too. Something else must have changed, and perhaps in a way that interacted with public policies regarding interest rates and home ownership.
That “something else” can be broadly characterized as “optimism” (perhaps rational, perhaps not) about the future because the housing boom featured rising housing prices and low net operating surplus. The facts that the housing boom ended with a large stock of housing yet real housing prices did not fall to pre-boom levels suggests that some of that optimism still remains in the marketplace.
One achievement of this paper is to quantify the expected size of the shift in the long run demand for housing structures that would be consistent with any given peak housing price elevation. For example, Figure 7 reports that real housing prices can be elevated more than 40 percent if the derived demand for housing structures were soon expected to shift by 0.40 log points in the quantity dimension, or 0.46 log points if that shift were anticipated 5 years in advance. To some readers, the stock of housing cannot be reasonably expected to increase that much, even in the long run, so that this result by itself confirms their suspicion that the housing cycle was indeed a “bubble”: a housing boom that was divorced from reasonable expectations about the economic fundamentals.
If the housing cycle is to be linked to fundamentals, then it must be explained why housing structures’ demand would shift so much. Another contribution of this paper is to show how, even in the absence of financial leverage, housing structures are a naturally and significantly leveraged claim on the value of housing services because some of the payments for those services must be used for depreciation, spending on intermediate goods, and property taxes, with only the residual remaining for the owners of structures. Thus, an expected 40 percent increase in real housing service expenditure could elevate housing prices by more than 40 percent in the short run.
Perhaps housing demand was expected to increase as a growing fraction of the population chose to work from home, entertain at home, and otherwise demand more housing. But this paper also points out that reduced production costs in the mortgage and real estate sectors – “housing intermediates” – could ultimately make existing housing more productive. Table 1’s first and second columns show that the derived demand for structures is significantly reduced by the necessity of spending resources on housing intermediates, and thereby that the long run housing stock would be significantly increased if some of those costs could be eliminated. I showed, for example, that a 2/3 probability of a modest housing service demand increase combined with a 75 percent reduction in the costs of housing intermediates would elevate prices by 40 percent or more. An important step for future research is to determine whether such expectations were quantitatively reasonable, and to monitor actual technical progress in the housing intermediates sectors.
The prospects for technical progress are inherently uncertain, so that their effects might be magnified by “irrational exuberance,” subsidies, or both. For example, market participants may focus too much on the rosy technology scenario in which technical progress is realized rather quickly, or in an especially large magnitude. Implicit government guarantees and reasonable prospects for government bailouts may also cause market participants to rationally downweight the less favorable states of nature.
Flood and Hodrick (1990, p. 99) found that “when agents expect the future to be somewhat different than history” price movements induced by fundamentals may look like bubbles to a researcher who is not fully aware of those fundamentals. In this regard, it is difficult to distinguish bubbles from rational expectations, and thus difficult to know whether the housing cycle of the 2000s is something that should have been “fixed.” However, I have strengthened this application of rational expectations theory to assume that agents’ expectations for progress are at least partly correct ex poste – that is, that some progress actually does occur (even if it were smaller or later than originally anticipated) – in which case the two theories are different in terms of the housing stock path after the bust. Perhaps the recession and transitory credit market interruptions are obstacles to immediate execution of this test, but one day we will know whether the housing stock (relative to trend) will ultimately remain above what it was before the housing boom, or instead that the construction boom of the 2000s was completely unnecessary.
The prospects for technical progress are exciting, but they are inherently uncertain. We do not yet know how to cure cancer, or how to make a cheap personal computer run at 100 Ghz. Nevertheless, we would be ill-advised to make investment decisions today under the assumption that lifetimes will remain constant, or that personal computers will always run at the same speed. Some of the inputs that are complementary with future technologies – those that are inelastically supplied in the short run – should begin to be accumulated today, even though the progress on which they rely has not yet occurred.
If the new technologies are realized earlier – or in greater magnitude – than originally anticipated, then the complementary inputs will enjoy a favorable and abnormal return. If the new technologies are realized later, or in lesser magnitude, then the complementary inputs will be under-utilized for a while and thereby earn an unfavorable return. This paper modifies the existing economic theory of housing investment to include these ideas. I conclude that at least part of the housing boom occurred because technology was expected to advance in ways that would make existing housing structures more productive.
When a bust occurs as a consequence of bringing the future into better (and less exuberant) focus, the boom time construction was, in hindsight, too much and too soon, but nonetheless the housing stock might not return to its pre-boom trend. As long as there was at least partial truth in the original anticipations – i.e., at least some productivity advance really did occur – then part of the construction boom was necessary and efficient, even with the benefit of hindsight. Housing prices must crash during the bust, but net housing investment may continue to be positive during the bust period. A “bubble” – that is, an increase in housing prices that occurred for no fundamental reason – would not have these properties. Thus, a bust period that keeps the housing stock and housing prices above their pre-boom trends means that housing fundamentals are in fact better, and thereby rejects the pure bubble theory.
Figure 1 shows the basic price and investment patterns of housing boom and bust to be explained. The left scale shows two monthly measures of the average value of housing properties, known as the “Case-Shiller Home Price Index” and the “OFHEO Home Price Index” divided by the PCE deflator and normalized to 100 in the first month shown (January 2000 [note seasonal adjustment]). The right scale shows residential construction spending, expressed as a percentage of personal income. The bottom series measures tenant-occupied housing construction and the top series adds owner-occupied construction. The Figure shows that real residential construction grew as housing values did through 2005. Sometime between late 2005 and early 2006, both values and construction spending peaked and then declined through early or mid 2009. Since then, construction spending increased a bit, as did one of the measures of real home values (both value measures agree that the 2009 change was significantly less than the changes during 2007 and 2008). The Figure also shows that the large majority of the amount and changes in residential construction spending are for owner-occupied structures.
Figure 2 displays quantity indices (2000 = 100) for investment in three types of structures: owner-occupied residential, tenant-occupied residential, and non-residential. The three types of structures investment stayed in about the same proportions 1993-2000. Both types of residential investment increased significantly 2001-2005, and thereafter fell back to pre-boom amounts. In contrast, investment in non-residential structures fell after 2000, and increased significantly 2005-2008. Because the boom increased both types of residential investment and reduced non-residential structures investment, the question addressed in this paper is why housing prices and construction increased 2000-2005, without particular attention to the distinction between owner-occupied and tenant-occupied housing.
The time patterns for housing prices, investment (both residential and non-residential), and income flows qualitatively match my stochastic technology model, and are unlike market responses to a short run shift in housing demand. Thus, the bulk of the housing boom appears to be a response to expectations about the future, and the housing bust a revision of those expectations. Some of those expectations are related to technical change, although it is also possible that additional expectations related to the anticipation of government bailouts, and even irrational exuberance, added to – or even multiplied – the effects of prospects for technical progress.
Section I explains how much of housing output goes for banking services, real estate brokerage, and management, as opposed to the owner of the structure. Structures have a leveraged claim on housing output because they are less elastically supplied than are the housing sector’s other inputs. Moreover, prospects for technical progress in the provision of the elastically supplied inputs should increase both housing prices and housing construction, especially for types of houses that are relatively intensive in the elastically supplied inputs. Section II amends the familiar “q-theory” model of housing investment with an elastically supplied intermediate goods sector. I measure time series for the net operating surplus and vacancy rates emphasized by the theory, and show how they suggest that optimism about the future fueled the housing boom, rather than tastes, technologies, or subsidies during the boom. Section III therefore looks at the market dynamics that would be an efficient response to good news about the future. It shows the expected size of the housing service demand shift – that is, the amount that the housing stock is expected to increase in the long run – that is implied by a given amount of short run housing price elevation. It shows the quantitative expectations about tastes and technology that are required to generate large housing price increases in the short run and housing stock increases in the long run, and how to distinguish the theory from alternative theories of the housing cycle, leaving future research to answer the question of whether such expectations were reasonable. Section IV points out that real housing prices are still higher than they were before the housing boom, despite the boom’s apparent legacy of overbuilding, which suggests that some of the boom’s optimism still remains three years later.
The housing boom that lasted through 2006 featured high and rising real housing prices, high rates of real investment in both owner-occupied and tenant-occupied housing, and low rates of real investment in non-residential structures. Housing sector net operating surplus – that minority part of housing service rents that accrues to homeowners and their creditors – was low relative to the housing stock during the boom, as were occupancy rates. More recently, real housing prices are much lower, although probably still above their pre-boom levels. Housing investment is now below, and non-residential structures investment above, pre-boom levels.
It is sometimes suggested that the housing boom derived largely from low interest rates, and public policies promoting homeownership, in the mid 2000s. But low interest rates by themselves should push up all kinds of structures investment, not just residential structures. And policies to encourage homeownership should have depressed construction of rental housing units, when in fact that construction boomed too. Something else must have changed, and perhaps in a way that interacted with public policies regarding interest rates and home ownership.
That “something else” can be broadly characterized as “optimism” (perhaps rational, perhaps not) about the future because the housing boom featured rising housing prices and low net operating surplus. The facts that the housing boom ended with a large stock of housing yet real housing prices did not fall to pre-boom levels suggests that some of that optimism still remains in the marketplace.
One achievement of this paper is to quantify the expected size of the shift in the long run demand for housing structures that would be consistent with any given peak housing price elevation. For example, Figure 7 reports that real housing prices can be elevated more than 40 percent if the derived demand for housing structures were soon expected to shift by 0.40 log points in the quantity dimension, or 0.46 log points if that shift were anticipated 5 years in advance. To some readers, the stock of housing cannot be reasonably expected to increase that much, even in the long run, so that this result by itself confirms their suspicion that the housing cycle was indeed a “bubble”: a housing boom that was divorced from reasonable expectations about the economic fundamentals.
If the housing cycle is to be linked to fundamentals, then it must be explained why housing structures’ demand would shift so much. Another contribution of this paper is to show how, even in the absence of financial leverage, housing structures are a naturally and significantly leveraged claim on the value of housing services because some of the payments for those services must be used for depreciation, spending on intermediate goods, and property taxes, with only the residual remaining for the owners of structures. Thus, an expected 40 percent increase in real housing service expenditure could elevate housing prices by more than 40 percent in the short run.
Perhaps housing demand was expected to increase as a growing fraction of the population chose to work from home, entertain at home, and otherwise demand more housing. But this paper also points out that reduced production costs in the mortgage and real estate sectors – “housing intermediates” – could ultimately make existing housing more productive. Table 1’s first and second columns show that the derived demand for structures is significantly reduced by the necessity of spending resources on housing intermediates, and thereby that the long run housing stock would be significantly increased if some of those costs could be eliminated. I showed, for example, that a 2/3 probability of a modest housing service demand increase combined with a 75 percent reduction in the costs of housing intermediates would elevate prices by 40 percent or more. An important step for future research is to determine whether such expectations were quantitatively reasonable, and to monitor actual technical progress in the housing intermediates sectors.
The prospects for technical progress are inherently uncertain, so that their effects might be magnified by “irrational exuberance,” subsidies, or both. For example, market participants may focus too much on the rosy technology scenario in which technical progress is realized rather quickly, or in an especially large magnitude. Implicit government guarantees and reasonable prospects for government bailouts may also cause market participants to rationally downweight the less favorable states of nature.
Flood and Hodrick (1990, p. 99) found that “when agents expect the future to be somewhat different than history” price movements induced by fundamentals may look like bubbles to a researcher who is not fully aware of those fundamentals. In this regard, it is difficult to distinguish bubbles from rational expectations, and thus difficult to know whether the housing cycle of the 2000s is something that should have been “fixed.” However, I have strengthened this application of rational expectations theory to assume that agents’ expectations for progress are at least partly correct ex poste – that is, that some progress actually does occur (even if it were smaller or later than originally anticipated) – in which case the two theories are different in terms of the housing stock path after the bust. Perhaps the recession and transitory credit market interruptions are obstacles to immediate execution of this test, but one day we will know whether the housing stock (relative to trend) will ultimately remain above what it was before the housing boom, or instead that the construction boom of the 2000s was completely unnecessary.
What Does Korea's Development Have in Common With the UK & the US?
Find out the answer, based on supply and demand, in Marc Teignier-Baque's The Role of Trade in Structural Transformation. He just earned his PhD at the University of Chicago (the only one wearing the tie)!
Are Real Housing Prices Still High?
Yesterday I showed how the Census Bureau New Home Price Index, adjusted for inflation, was no lower -- probably higher -- now than it was throughout the 1990s.
Remarkably, several bloggers attack the piece on the basis that the Census Bureau index is supposedly not quality adjusted. That's a funny claim, because the Census Bureau refers to the index as the "Constant Quality (Laspeyres) Price Index of New One-Family Houses Sold."
I clearly explained that I chose that index because it dates back far enough to examine other housing cycles (e.g., those in the 1970s and early 1980s), which gives me a historical basis to model how supply and demand would react in various hypothetical situations.
But let's suppose that one did not want to engage in such an exercise, and thereby had the freedom to use other indices: my conclusion from yesterday -- that real housing prices are still high by historical standards -- would be even stronger!
To see this, compare the end of 2009 with the last quarter of 1999 or the first quarter of 2000 (that's as far back as some of the alternative indices go), as I have in Table 3. The last column of the Table has the index I used yesterday, and it shows the LOWEST 2009 real housing price relative to 10 years ago, with one exception. Eight other measures show MORE of an increase over 10 years.
The various national measures disagree much more on the amount by which real housing prices increased 2000-2006, with a range from 20% (Census Bureau quality-adjusted new homes) to 73% (Radar-Logic). A significant part of the disagreement relates to the weighting of the various regions to form the national composite, which is why my Table 3 also reports unweighted averages of the regions, and real price changes for the median region.
Remarkably, several bloggers attack the piece on the basis that the Census Bureau index is supposedly not quality adjusted. That's a funny claim, because the Census Bureau refers to the index as the "Constant Quality (Laspeyres) Price Index of New One-Family Houses Sold."
I clearly explained that I chose that index because it dates back far enough to examine other housing cycles (e.g., those in the 1970s and early 1980s), which gives me a historical basis to model how supply and demand would react in various hypothetical situations.
But let's suppose that one did not want to engage in such an exercise, and thereby had the freedom to use other indices: my conclusion from yesterday -- that real housing prices are still high by historical standards -- would be even stronger!
To see this, compare the end of 2009 with the last quarter of 1999 or the first quarter of 2000 (that's as far back as some of the alternative indices go), as I have in Table 3. The last column of the Table has the index I used yesterday, and it shows the LOWEST 2009 real housing price relative to 10 years ago, with one exception. Eight other measures show MORE of an increase over 10 years.
The various national measures disagree much more on the amount by which real housing prices increased 2000-2006, with a range from 20% (Census Bureau quality-adjusted new homes) to 73% (Radar-Logic). A significant part of the disagreement relates to the weighting of the various regions to form the national composite, which is why my Table 3 also reports unweighted averages of the regions, and real price changes for the median region.
Wednesday, May 5, 2010
Was It Really A Bubble?
Copyright, The New York Times Company
Adjusted for inflation, residential property values were still higher at the end of 2009 than 10 years ago. This fact raises the possibility that at least part of the housing boom was an efficient response to market fundamentals.
Inflation-adjusted housing prices and housing construction boomed from 2000 to 2006 and crashed thereafter. Commentators ranging from President Obama to Federal Reserve Chairman Ben S. Bernanke have described that cycle as a “bubble,” by which they mean that, at least in hindsight, the housing price boom was divorced from market fundamentals.
But maybe there was a good, rational reason for housing prices to increase over the last decade.
Let’s consider first what it means to believe that the spike in prices since the late 1990s was unwarranted — the so-called “bubble theory.”
According to the bubble theory, for a while the market was overcome with exuberance, meaning that people were paying much more for housing than changes in incomes, demographics, technology and other basic factors would suggest.
Now that the bubble is behind us, people today should be no more willing to pay to own a house than they were in the late 1990s. (It’s true that population has grown since the 1990s, but population growth is nothing new and should not by itself increase real housing prices. Don’t forget that greater population also means more people available to do construction work.)
Bubble theory also implies that the technology for building homes and providing housing is not very different today from what it was in the late 1990s, and so you couldn’t blame the spike in prices on changing building technologies either. In other words, the bubble theory blames rise in prices on animal spirits, and not on any “legitimate” increase in building costs.
Meanwhile, bubble theorists also say that today America is “overbuilt” as a result of the bubble. With too much housing and no additional demand to be supported by market fundamentals, real housing prices should, according to the bubble theory, be lower today than they were in the late 1990s.
A reasonable estimate, based on bubble theory, is that housing inventory is about 3 or 4 percent above what it would have been without the bubble and without the temptation to overbuild. In order to make these excess homes worth buying, prices need to fall further; economists would generally estimate that an extra 1 percentage point of housing inventory requires a matching 1 percentage point decline in price to make those excess homes look like a good deal.
So if we believe we had 3 or 4 percent more homes than we really needed last year, based on market fundamentals, that means that housing prices would eventually be about 3 or 4 percent below what they were before the bubble, in order to make those extra houses worth purchasing.
I’ve plotted this path out in the blue series in the chart below. It shows the housing price path one might have expected for 2009 and beyond if housing demand had returned to about where it was before the housing boom.
A note: Here housing prices are measured according to the Census Bureau index, which dates back far enough to observe previous cycles, and adjusted for inflation using the Bureau of Economic Analysis’s price index for personal consumption expenditures. The index is normalized so that it is 100 in 1994-97. In other words, a value of “100” for today would mean that inflation-adjusted new home prices are no different today than they were then.
According to the blue series, created using the “bubble theory” premises, real housing prices would be 3 or 4 percent lower now than they were before the housing boom (a value of about 97 for the index) if in fact housing demand were no different, because the “overbuilt” inventory of houses is supposedly 3 or 4 percent greater now than it was then.
This path approximates what some bubble theorists were expecting last year. For example, during the first half of 2009 Professor Robert Shiller (the same Shiller from the Case-Shiller housing price index) forecasted that housing prices would fall further.
Of course, our population continues to grow, so the housing boom’s excess inventory will not last forever as new households form. But this process takes a while, which is why the blue series in the chart shows real housing prices returning to normal (a value of 100 for the index) only slowly after 2010.
Now let’s turn to the black series in the figure, which shows actual housing prices.
Contrary to the blue “bubble theory” series, actual housing prices have risen slightly over the last four quarters, so far remaining well above what they were before the housing boom. Although real housing prices are sharply off their highs, today they appear high by historical standards.
Bubble theorists might say that some unfounded optimism lingers in the market place, and that Professor Shiller was just a little early in predicting that real housing prices will soon be significantly below what they are now.
But another interpretation is that a large fraction of the housing price boom was justified by fundamentals (and next week I’ll consider some of the specific fundamentals that may have permanently increased housing demand in the 2000s). If so, we are probably asking too much of the Federal Reserve and other regulators to accurately disentangle bubbles from fundamentals the next time that asset prices rise.
Adjusted for inflation, residential property values were still higher at the end of 2009 than 10 years ago. This fact raises the possibility that at least part of the housing boom was an efficient response to market fundamentals.
Inflation-adjusted housing prices and housing construction boomed from 2000 to 2006 and crashed thereafter. Commentators ranging from President Obama to Federal Reserve Chairman Ben S. Bernanke have described that cycle as a “bubble,” by which they mean that, at least in hindsight, the housing price boom was divorced from market fundamentals.
But maybe there was a good, rational reason for housing prices to increase over the last decade.
Let’s consider first what it means to believe that the spike in prices since the late 1990s was unwarranted — the so-called “bubble theory.”
According to the bubble theory, for a while the market was overcome with exuberance, meaning that people were paying much more for housing than changes in incomes, demographics, technology and other basic factors would suggest.
Now that the bubble is behind us, people today should be no more willing to pay to own a house than they were in the late 1990s. (It’s true that population has grown since the 1990s, but population growth is nothing new and should not by itself increase real housing prices. Don’t forget that greater population also means more people available to do construction work.)
Bubble theory also implies that the technology for building homes and providing housing is not very different today from what it was in the late 1990s, and so you couldn’t blame the spike in prices on changing building technologies either. In other words, the bubble theory blames rise in prices on animal spirits, and not on any “legitimate” increase in building costs.
Meanwhile, bubble theorists also say that today America is “overbuilt” as a result of the bubble. With too much housing and no additional demand to be supported by market fundamentals, real housing prices should, according to the bubble theory, be lower today than they were in the late 1990s.
A reasonable estimate, based on bubble theory, is that housing inventory is about 3 or 4 percent above what it would have been without the bubble and without the temptation to overbuild. In order to make these excess homes worth buying, prices need to fall further; economists would generally estimate that an extra 1 percentage point of housing inventory requires a matching 1 percentage point decline in price to make those excess homes look like a good deal.
So if we believe we had 3 or 4 percent more homes than we really needed last year, based on market fundamentals, that means that housing prices would eventually be about 3 or 4 percent below what they were before the bubble, in order to make those extra houses worth purchasing.
I’ve plotted this path out in the blue series in the chart below. It shows the housing price path one might have expected for 2009 and beyond if housing demand had returned to about where it was before the housing boom.
A note: Here housing prices are measured according to the Census Bureau index, which dates back far enough to observe previous cycles, and adjusted for inflation using the Bureau of Economic Analysis’s price index for personal consumption expenditures. The index is normalized so that it is 100 in 1994-97. In other words, a value of “100” for today would mean that inflation-adjusted new home prices are no different today than they were then.
According to the blue series, created using the “bubble theory” premises, real housing prices would be 3 or 4 percent lower now than they were before the housing boom (a value of about 97 for the index) if in fact housing demand were no different, because the “overbuilt” inventory of houses is supposedly 3 or 4 percent greater now than it was then.
This path approximates what some bubble theorists were expecting last year. For example, during the first half of 2009 Professor Robert Shiller (the same Shiller from the Case-Shiller housing price index) forecasted that housing prices would fall further.
Of course, our population continues to grow, so the housing boom’s excess inventory will not last forever as new households form. But this process takes a while, which is why the blue series in the chart shows real housing prices returning to normal (a value of 100 for the index) only slowly after 2010.
Now let’s turn to the black series in the figure, which shows actual housing prices.
Contrary to the blue “bubble theory” series, actual housing prices have risen slightly over the last four quarters, so far remaining well above what they were before the housing boom. Although real housing prices are sharply off their highs, today they appear high by historical standards.
Bubble theorists might say that some unfounded optimism lingers in the market place, and that Professor Shiller was just a little early in predicting that real housing prices will soon be significantly below what they are now.
But another interpretation is that a large fraction of the housing price boom was justified by fundamentals (and next week I’ll consider some of the specific fundamentals that may have permanently increased housing demand in the 2000s). If so, we are probably asking too much of the Federal Reserve and other regulators to accurately disentangle bubbles from fundamentals the next time that asset prices rise.
Monday, May 3, 2010
Labor Market Distortions in Mexico Since 1980
Economist Gabriel Martinez has studied the evolution of the Mexican labor market, finding that labor market distortions there decreased significantly in the 1980s and 1990s, much of which may be associated from changes in tax laws.
See especially his Figure 16.
See especially his Figure 16.
How Big was the Housing Price Cycle?
Various housing "price" indices agree than the 2000s had a large housing cycle, with real housing values rising over a prolonged period, and then falling back close to their initial values in just a year or two. But they do not agree on the magnitude of the price boom, as shown in Table 1. The Table’s ten measures differ in terms of whether they compare same properties over time (Case-Shiller, Radar Logic, and OFHEO), or hedonically adjust new housing prices for measured characteristics (Census Bureau). Table 1 reveals how much of the difference among the same-property indices is related to their aggregation of the various cities or MSAs sampled, because the range of medians is only 10 percentage points, as compared to 38 percentage points for the national composites.
The Case-Shiller and Radar-Logic are commercially sponsored indices: it is possible that the MSAs sampled are not random with respect to the elasticity of housing supply (the sampled MSAs have less elastic supply than the un-sampled). Consistent with this hypothesis, the Case-Shiller composite-10 (the 10 MSAs they have followed longer) appreciates more than the Case-Shiller composite 20. For more research on reconciling various housing price indices, see some of these posts and the references cited therein:
C-S/OFHEO comparison from Professor Nunes
A New Comparative Analysis of Housing Price Indices
Case-Shiller OFHEO faceoff
Regional shifts may cause housing prices to rise in some places and fall in others, so I take 40-50% as the size of the housing property value boom to be explained by national factors such as exuberance, monetary policy, or technical change. Moreover, "housing property values" are not the same as "housing prices" because the former also reflects changes in the quantity of housing on a given property and changes in land prices. Conceptually, the Census Bureau's quality-adjusted index of new home prices could well be the best indicator of housing prices, and it shows a 20 percent housing price peak.
I think it's safe to say that, from a national perspective, the housing price cycle involve peak housing price appreciation of 20-40 percent.
The Case-Shiller and Radar-Logic are commercially sponsored indices: it is possible that the MSAs sampled are not random with respect to the elasticity of housing supply (the sampled MSAs have less elastic supply than the un-sampled). Consistent with this hypothesis, the Case-Shiller composite-10 (the 10 MSAs they have followed longer) appreciates more than the Case-Shiller composite 20. For more research on reconciling various housing price indices, see some of these posts and the references cited therein:
C-S/OFHEO comparison from Professor Nunes
A New Comparative Analysis of Housing Price Indices
Case-Shiller OFHEO faceoff
Regional shifts may cause housing prices to rise in some places and fall in others, so I take 40-50% as the size of the housing property value boom to be explained by national factors such as exuberance, monetary policy, or technical change. Moreover, "housing property values" are not the same as "housing prices" because the former also reflects changes in the quantity of housing on a given property and changes in land prices. Conceptually, the Census Bureau's quality-adjusted index of new home prices could well be the best indicator of housing prices, and it shows a 20 percent housing price peak.
I think it's safe to say that, from a national perspective, the housing price cycle involve peak housing price appreciation of 20-40 percent.
Subscribe to:
Posts (Atom)
