September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« Apparently Environmentalists Have Different Definitions of "Credible" and "Authoritative" | Main | Law, Economics, and Superstition Part 3: "Gypsies" »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.


Why wouldn't people rent housing in the new location, and rent out their old home while they wait (and hope) for its value to rise?

I guess rents would often be higher in areas with better job opportunities, but income from the new job should more than cover the difference.

Of course unemployment benefits could play a role here.

If you have a house then you can enter the buying & selling market or the rental market.

Last time I moved for a job I kept my old property. I rented that place out and rented another place in the town I moved to. Why isn't that sort of thing possible in the US?

Renting is possible, but:

If you are stuck with a large mortgage payment, you may well be unlikely to charge enough rent to cover it. If the rent on the new place is greater than what you can get on the old place (and if it's not selling, it might not fetch much in rent), then you're actually ending up with a larger housing payment if you take this strategy.

If the worker in question is married with kids, finding a house to rent is not always easy. My sense is that the apartment rental market is more thick than the house rental market, and if you have 2 or 3 or more kids, finding a suitable apartment can be a challenge.

Add in the fixed costs of moving and it would make sense that people would at least *delay* any decision to look elsewhere for new work. Yes, unemployment is up but it's duration that's the interesting piece. The housing market problems suggest that people are just engaging in more search before moving because moving involves a bigger loss than it used to, thanks to the large mortgage and little equity.

Yes, David, unemployment benefits matter here and there is evidence that when they run out that people tend to find jobs fairly quickly. I didn't mean to suggest the housing piece was the only explanation for the increased duration, but that it might be one piece of the puzzle.

Why doesn't the recent federal minimumwage increase do a big part of the trick? (http://www.laborlawcenter.com/t-federal-minimum-wage.aspx) Combine that with the Obamacarebill which raises costs of employing more workers and I would say you have a rather fair explanation.

I'm not sure if I'm convinced by the 'they 'can't move argument', because in contemporary America people have a big area surrounding them in which they can find employment and get their in a reasonable time. (How big is the area you can try to find work in if you have, at most, an hour to drive to work?)

But, of course, I might be missing something.

Remember: the thing to be explained here is why people are unemployed for a longer duration. I'm not sure that minimum wage will get you there, although that might contribute to explaining the higher overall rate and its effects on various groups.

The point is this: if you are unemployed and you could move and possibly find a new job, why wouldn't you do it? If moving means higher housing costs (due to the loss on the home if you could sell it or the rent scenario above), then you are that much more likely to continue to search for a job in your current location.

It doesn't seem unreasonable to imagine people looking at the choice between job hunting here vs. job hunting somewhere else and saying "I can't afford to move, so I might as well take my chances here rather than there." In prior recessions, where the housing piece wasn't so salient, people might have more easily been willing and able to move on the possibility of a job. Less so this time, is the argument.

I wonder too if this isn't linked to the increased number of people leaving the labor force. Are some of those folks people who have just given up on searching "at home" and can't afford to take the loss on trying to move elsewhere?

Again, more conjectures to be refuted. :)

There is geographical sectional data on the unemployment rates and durations around the country.

Perhaps this could be compared to the housing markets in different parts of the country.

Richard Ebeling

I have heard Steve's point made in a somewhat different context. Housing is an illiquid asset, and more so now in many markets. What people did in boom times is not necessarily relevant.

To the degree people do move and rent their homes out, that would put downard pressure on rents. That would worsen the problem for those under water with their mortgages. (Query: what is happening in rental markets?)

To take a stab at Jerry's query, it seems that many rental markets have been fairly bouyant. The current (July 19-25) issue of Bloomberg Businessweek has an interesting article, "Renters to the Rescue in Miami." It's only one datapoint in a mixed market, but renting their home has perhaps been a solution for some people.

Steve's argument is sound, but isn't it at least slightly based on the illusion that the "problem" is that people have been laid off from their old jobs, and they aren't filling all of the vacancies that exist somewhere else?

This recession did have a jump in layoffs when money expenditures were dropping fast, but for the most part, there has been only a slight increase in layoffs and other separations, and a big drop in quits and a bigger drop in new hires. And the vacancy rate is much lower.

I doubt that many unemployed new college grads are being deterred from job opportunities because they need to sell their house, but I would like to advise my students where they should move? Where is this geographic region with the labor shortages? (I realize that most of the country has less of a problem than here in S.C.)

If you look at the demographics regarding the unemployment rate (though I haven't looked at the figures for duration,) the unemployment rate is much higher for younger people. Those that have bought houses are especially likely to be underwater, I suppose. But I suspect that the groups with especially high unemployment rates are also especially likely to be renters.

Wouldn't the constantly-expanding duration of unemployment insurance have an effect? Many people, if they are receiving more unemployment, may hold out for better jobs than they would otherwise get.

Also, just the economic uncertainty currently at play is likely keeping employers from hiring -- which would of course lengthen duration.

And even in a recession, I am willing to bet that businesses are still more willing to hire people who already have jobs over those who do not, no matter why they don't. So the few jobs opening up are likely to be going to the already-employed.

Prof. Horwitz:

Higher minimum wages tend to create long-term unemployment, it's not a short term phenomenon.

In Italy unemployment in the south is 20-30%, as among the youth.

These workers are submarginal at the legal labor cost (ther is no minimum wage, legally, in Italy, only compulsory collective bargaining and lots of taxes).

People whose human capital has been destroyed by the crisis (e.g., housing-related jobs with no economic usefulness at present), young people with no experience or training, etc... they could find a job at the inflated demand of the boom, they cannot do it now without a reduction in their nominal and real wages.

It appears that most of the new unemployed were on the left tail of the distribution: this shows why average earnings increased despite the increase in unemployment. These people are the most likely to suffer because of minimum wage laws, and the ones who are submarginal today will keep on being submarginal for a long time.

Pietro,

For your argument to explain why this recession has had notably longer median duration of unemployment (if we could hold unemployment benefits constant), it would have to be the case that the minimum wage TODAY is taking more of a "bite" out of the labor market than it did in past recessions, or that the workers who are unemployed in this recession are in groups particularly hit by the minimum wage.

I'm not at all convinced the first is true. In fact, a quick look at the data suggests it isn't true. Seat of the pants: I divided the minimum wage by the average private sector wage in the year of each recession dating back to 1973. The results are here:

1973-74: 44.6%
1981: 46.5%
1982: 43.4%
1991: 41.0%
2001: 36.0%
2009: 39.3%

So relative to the average private sector wage, the minimum wage is lower now than in all but one of the prior 5 recessions. Seems like the minimum wage has less bite than in prior recessions. So I'm not sure that answers the question of unemployment duration, if my logic is right here.

I'm also not sure this recession has hit industries affected by the MW particularly hard either.

Actually, I take back the last sentence: there is evidence that the unemployment rate is particularly high among the unskilled, so that might be some evidence that the MW is more of a problem than I first argued.

It is instructive to look at the international record, which is shocking. The US is clearly a laggard in jobs recovery. This is discussed in a letter in today's WSJ and here is a link to a July 10th WSJ article with charts.

http://online.wsj.com/article/SB10001424052748704799604575357031890309998.html?mod=ITP_pageone_1

Let's talk about real people -- in the real world, on the ground in some of the states with the highest unemployment rate.

1 out of 4 of the unemployed nation wide are construction workers. The numbers are higher in Califoria.

In California a large percentage of these are foreign nationals in the country illegally.

I've talked to some of these folks.

Mobility back to their homes means going back to Mexico or Latin America -- and they know family members and co-workers have indeed returned to their homes in their country.

The "mobility" problem in terms of homes here is making the choice of whether to choose to go on the other side of the border.

Greg -- Would the illegals be counted among the unemployed? Or the previously employed?

Jerry -- I don't know.

I've asked on another blog and got no answer.

Anyone know the answer?

Does it depend on what sort of illegal documents they are using?

The government is reporting huge unemployment numbers for border towns were the illegals live.

But I don't know anything about what happens on the ground when they do this counting (the "official" procedure is likely only slightly related to what the bureaucracy actually does).

Jerry writes,

"Greg -- Would the illegals be counted among the unemployed? Or the previously employed?"

To provide a little anecdotal evidence in support of the original theory: my brother just told me he and his wife had thought about selling their house, but when they found out what they could get for it -- well below what they paid, and what they owed -- they decided to stay put. And the DFW area hasn't been hit that hard by the housing crisis, either.

It seems to me that if unemployment continues for much longer, though, people will end up having to leave, meaning they will have to sell, and housing prices will drop quite a bit.

Reinhart and Rogoff exmined the history of serious financial crises like the current one. Housing prices are typically hit hard. On average, REAL housing prices decline for 6 years and fall 35.5%. If past is prologue, we're halfway through the decline at best.

several reasons I'm skeptical:

1. Long term unemployment and unemployment in general seems to be excessively high among the low skilled workforce in this recession. Low skilled workers tend not to be homeowners so wouldn't be hampered by underwater mortgages preventing them from selling their homes to move to better job climates.

1.1 To the degree that lower skilled workers did get homes in the big boom, they would tend to be lower priced homes. In general, low priced homes move up and down less dramatically than higher priced homes. This makes it less likely IMO that a low skilled worker would see their home suddenly turn deep underwater. I can see this as more of a factor among professionals who brought a home for, say, $750,000 and now discover the market will only fetch $500,000 for it leaving them the option of toughing it out or risking their credit in letting the bank grab it. The starter home brought for $100K has probably not dropped to $50K except in a few 'ghost town' like areas. It's probably more like $90K which means the 'underwater' issue isn't quite as dramatic.

2. The massive foreclosure numbers as well as homeowners struggling with underwater mortgages creates a fantastic renters market. That means for the underwater unemployed homeowner, the cost of leaving their home may have gone up but the cost of moving to a new place has gone down.

3. This assumes the recession has been driven by intense geographical forces, but I don't think the evidence really bears that out. Yes there are some areas that have suffered horribly from the housing bust (Nevada, Florida, but these areas have often seen housing booms and busts) but there's no real evidence or logic that says the recession indicates the need for dramatic geographical relocation of huge numbers of Americans. I don't see any evidence that a large and diverse economy such as California suddenly needs to relocate large numbers of its people to, say, Texas or Seattle. If the 'recalculation' aspects of this recession doesn't require great geographical relocation then unemployed workers 'trapped' in homes they can't sell needn't be a major factor.

How good is the system of policing welfare in the US? If you claim unemployment and do work on the side how likely are you to be caught?

Note well.

Hayek identifies housing production as a long-period production good for the purposes of his malinvestment through the production structure and credit / leverage structure account of the boom / bust.

"Reinhart and Rogoff exmined the history of serious financial crises like the current one. Housing prices are typically hit hard. On average, REAL housing prices decline for 6 years and fall 35.5%. If past is prologue, we're halfway through the decline at best."

Greg -- Do you have a citation for that? -- Jerry

Hayek identifies housing production as a long-period production good for the purposes of his malinvestment through the production structure and credit / leverage structure account of the boom / bust.

According to the census bureau the number of housing units in July 2000 was 116.3M. In July of 2009 it was 129.9M. Population over the same period went from 281M to 307M. In 2000 we had about 2.4 people per housing unit, in 2009 we had 2.36.

If by malinvestment you mean real, actual investment that is harmful in the long run as opposed to simply investment that does not earn as much return as original investors had hoped for, then the numbers do not fit. We did not 'overproduce' houses nor do we have some glut of too many houses in the US.

Can you give us a link Booton?

I find it interesting that you picked an end date during the bust. I'd be more interested in seeing what the data were in 2006 or 2007, and what exactly is being counted at housing units.

A comparison of housing starts might be more interesting. If new units were replacing old ones being torn down, the total number wouldn't change, but that sure as hell would be the sort of new capital investment that the Austrian theory predicts.

Housing units can be found at:

http://www.census.gov/popest/housing/ I didn't pick the years, just searched on "housing units by year in the us"

Not much pattern by year other than steady increase:

2000 116.3M
2001 117.9M
2002 119.45M
2003 121.1M
2004 122.8M
2005 124.7M
2006 126.47M
2007 128.1M
2008 129.3M
2009 129.9M

In terms of growth per year the highest is 2005 which was 1.5% larger than 2004. 2006 and 2007 were 1.4% and 1.3% 2009 was the lowest at 0.5%.

US population is at http://factfinder.census.gov/servlet/SAFFPopulation?_sse=on

In terms of malinvestment I'm not following your point about housing starts. OK what if we did tear down a lot of perfectly fine houses and build brand new ones in their place? Assuming the useful life of a house is 30-50 years we are still better off. We have houses that have fewer 'miles' on them than before. OK maybe we'd have been even better if we had put some of that investment in medical information processing, solar panels, cell phone tech or whatever else is out there but in terms of harm, there should be none. In real terms we are an economy with more capital than earlier therefore our productive capacity should be higher. Maybe it means some yuppie flippers go bankrupt while some artist types move into empty "McMansions" in empty mega-developments but there's no particular reason that should be a bad thing in and of itself.

Additionally, population growth was about 1.1% in the 1990 and is expected to decline very slowly to 0.54% around 2040-50. That would make a yearly increase in terms of units of about 1.5% not that outrageous.

Housing starts can be found at http://www.nahb.org/generic.aspx?genericContentID=554

2000 1.57M
2001 1.60M
2002 1.71M
2003 1.85M
2004 1.96M
2005 2.07M
2006 1.80M
2007 1.36M
2008 0.91M
2009 0.55M

The number seems to be between 1-2% of the total # of housing units. (Although this now gets confusing since multi-family houses would mean more than one unit). If you consider the useful life of a house to be about 50 years or so then it hardly seems unreasonable to replace about 1% of units every year simply due to depreciation and another 1% due to population growth. On a macro level I'm not seeing a glut of houses yet.

See _TPTofC_ p. 77 of the Collected Works edition, where Hayek writes,

"Included under this term [i.e. "capital"] are not only the man-made productive equipment insofar as it is not expected to remain useful forever but also natural resources insofar as they are 'wasting assets', and all consumers' goods existing at the moment insofar as they are non-permanent sources of final income .. fn15 It may sound courious that we reckon, e.g. houses as capital only if and insofar as they are non-permanent .. "

Check the reference for the full quote.

"Greg -- Do you have a citation for that? -- Jerry"

We're in the process right know of making repairs to a home about 8 years old, in order to keep it in shape -- the parts of a home do wear out as they age, and a home would rot and would be wrecked by vandals if there was not a constant process of oversight and maintenance.

Booton, are you a troll or do you genuinely not get this?

"f by malinvestment you mean real, actual investment that is harmful in the long run as opposed to simply investment that does not earn as much return as original investors had hoped for, then the numbers do not fit."

If you genuinely don't get this, I'd recommend figuring out what Hayek's trade cycle theory is about before we pursue this special topic any further.

Yes, Mr. Boonton needs to do a little reading to understand what Austrians mean by mal-investment. Some Austrian capital theory is in order, I'd say.

This sentence is particularly problematic: "In real terms we are an economy with more capital than earlier therefore our productive capacity should be higher."

Actually, no it shouldn't necessarily. The idea of "more capital" as if capital were a big glob of goo (or "play-dough" in Pete B's analogy) is precisely what Austrians reject. The question isn't "how much" capital but whether the capital we are producing "fits" with both the existing capital and consumer preferences. Capital is heterogenous: "more" capital isn't necessarily better if the "more" is stuff that isn't complementary to what the entrepreneur already is making use of.

Making more houses (or whatever) due to an artificially low interest rate means we have more physical stuff, yes. But for that stuff to be value-producing capital, it has to fit into other plans and be consistent with underlying time preferences. The bust we have been suffering, and the folks having mortgage issues, suggests it wasn't.

Once one talks in terms of "more capital" without taking heterogeneity and complementarity into account, one is missing the Austrian point.

BTW, more on this in Thursday's Freeman column.

Steve,

The numbers do not seem to imply we have an exceptional number of 'new houses' due to the boom. I also understand that consumer preferences matter here but aside from some 'ghost towns' were hundreds of houses sit vacant, we haven't all suddenly decided living in houses isn't cool anymore.

The evidence seems to suggest that while the boom caused a lot of lost money and bad loans, we still do not have any mass glut of 'too many' houses. We have about one house for every two and a half people which is about what we had before the boom. If, as you suggested, many supposedly good houses were torn down and brand new ones built in their place....well then we have on average one newer home for every two and a half people.

In itself that seems like a lot of useful capital that will be able to provide a type of income (places to live) for many years to come. Unlike, say, the thousands of miles of 'dark fiber' that was laid during the dot com boom....this capital is quite versatile (a house can provide a place to live for a doctor, a carpenter, a dentist, a school teacher, a data analyst, etc. there's not much you can do with fiber optic cable besides using it for digital communications). So I'm not really seeing the evidence to support the implication that we are suffering from a mass 'malinvestment' of produced homes.

Likewise I'm skeptical of the claim that unemployment is being held high by workers trapped in underwater homes for reasons I provided earlier.

How is "housing unit" being defined? Techinically, an apartment is a "housing unit." If those are being counted, then that could account for the apparent lack of overproduction in "housing units," since if actual houses are being built and people are moving into them, fewer apartments will be built.

Also, the whole subprime mortgage problem arose precisely because low skilled workers (or people not working at all) were buying houses. This was in addition to a few local housing booms -- Florida and Claifornia -- where the malinvestments were worst. A modest increase in housing overall masks these local booms. Be careful with statistics.

Troy,

First subprime != poor. Subprime means a high risk loan. A doctor swinging a $1M mansion with no down payment could just as much be a subprime loan as a plumber with a shoddy credit history buying a $150K starter home.

Second, we still have one 'housing unit' for every two and a half people. If lots of 'units' were upgraded from apartments to homes then as a result of this malinvestment we went from one low quality apartment 'housing unit' for every two and a half people to one new house 'housing unit' for every two and a half people. (But I don't think you're going to find lots of evidence for massive tear downs of apartment complexes during the boom, the boom wasn't so much about building lots of new homes as it was about paying higher and higher prices for homes....new or not).

I agree there were local housing booms that have produced 'ghost towns', although even this doesn't seem like it should be malinvestment. Sure its malinvestment for the developers and homeowners who borrowed to build or buy those homes, but after the dust settles many of them are perfectly good homes that can be brought or rented cheaply. Just as 'starving artists' moved into empty industrial lofts in many big cities and 'colonized' them, there's no particular reason something like that can't play out here unless the 'ghost towns' are so far away from civilization that there's nothing useful that can be done with the houses other than let them rot in place until they turn to dust.

That case aside, homes are still a piece of useful capital that should provide positive income for many years to come. While it makes sense that individual investors will suffer for years for their bad choices and bad luck, it's not clear to me how the homes should be some type of macro-millstone that must be tied around the economy's neck for years to come. In general more capital is more production unless the capital is so totally useless that the only efficient thing to do with it is let it rot. (BUT keep in mind that the numbers do NOT show either a dramatic increase in the # of homes nor even in terms of housing starts supposedly replacing existing homes. Relative to population and the natural depreciation of existing homes, the last ten years does not appear that out of line in terms of changes to the actual housing stock.)

Greg,

Boonton is a troll.

Greg,

You'll have to define troll for me here. The traditional definition is someone who feigns ignorance to bait others to post long explanations of concepts he already understands. A more malicious definition might be someone who baits others in an argument with only the intent of stirring hostility. I don't think either applies here. But perhaps you mean someone who disagrees with you. In that case your life on the Internet is not going to be very fun.

While I'll yield to others here who have greater knowledge of Austrian trade cycles but from what I've read of 'malinvestment' it appears to mean the popping of a bubble reveals that investments that were once thought productive were really not.

I do not see the evidence that the housing boom fits the data. For one thing, it's clear we didn't have an exceptional amount of house building. For another thing, a house is a remarkably useful piece of capital. Just about anyone can live in a house, they can easily be highly customized to individual tastes either on a budget or with a large decorating account, you can use them for storage, to rent out to people, heck for a lot of thing. They also take a very long time to depreciate.

Compare this to the 'malinvestment' of the dot-com bust. Fiber optics, web servers etc. have very limited uses and depreciate very rapidly. Yes a lot of capital from the dot-com bust was sold at a discount and put to other productive uses (the pets.com puppet now sells car loans, 'dark cable' lead to cheap long distance rates) but still there's only so much you can do with such capital. A house, on the other hand, is almost always useful in some way unless it's located so far away from civilization that no one would want to live in it.

"While I'll yield to others here who have greater knowledge of Austrian trade cycles but from what I've read of 'malinvestment' it appears to mean the popping of a bubble reveals that investments that were once thought productive were really not.

I do not see the evidence that the housing boom fits the data."

Well, you're an idiot then.

Another possibility is that you don't understand basic economic concepts.

Maybe both?

Greg,

We can do without the name calling please. Show him where he's wrong, but leave the other stuff on the middle school playground please.

Sad to see that the unemployment issue has been arising rapidly. The government should see this problems and utilize its resources in terms of dealing with the resolution.

The comments to this entry are closed.

Our Books