Bob Murphy wrongly assumes I won’t peek

I recently made this argument:

Yes, housing output was low in 2009 and unemployment was high.  But is there a causal relationship?  I say no.  Housing starts peaked in January 2006, and then fell steadily for years:

January 2006 — housing starts = 2.303 million, unemployment = 4.7%

April 2008 — housing starts = 1.008 million, unemployment = 4.9%

October 2009 — housing starts = 527,000, unemployment = 10.1%

So housing starts fall by 1.3 million over 27 months, and unemployment hardly changes.  Looks like those construction workers found other jobs, which is what is supposed to happen if the Fed keeps NGDP growing at a slow but steady rate.  Then NGDP plummeted, and housing fell another 480,000. 

Bob Murphy responded:

The reason is that there’s more to the construction sector than simply starting new houses. That fact alone buys us another few months, as a look at housing completions shows.  [he has a link here]

But more important,

Waaaait a minute.  Note how Bob tells me housing starts are the wrong data, because construction workers keep working for some time after the starts, and then tells us to look at housing completions.  But then he merely provides a link, moving right along to something “more important.”  I wonder why?  Perhaps because housing completion data also supports my view?  Here are the numbers.  I’ve also averaged the two, as the average of starts and completions might be a good indicator of ongoing activity.

January 2006:  starts 2,303,000   completions 2,058,000    average  2,180,000

April 2008:   starts   1,008,000    completions   1,014,000     average   1,011,000

October 2009   starts    527,000     completions   745,000    average   636,000

Using the housing activity average, an even greater share of the total slowdown occurred between 2006 and 2008, when unemployment was stable, and an even smaller share occurred after April 2008.  I want to thank Bob Murphy for further strengthening my argument.

Bob also makes another argument, citing data showing that construction employment declined much less than housing construction between 2006-08.  But that’s easy to explain, as commercial real estate prices didn’t peak until late 2008.  So the commercial RE sector may have picked up some of the workers laid off from building houses.   (I don’t know about infrastructure and government building.)  And even if commercial RE didn’t add housing workers, if housing is half of all construction then a 20% decline in housing construction jobs would translate into only a 10% decline in all construction jobs.  All this of course supports my point.  The big drop in housing construction between January 2006 and April 2008 did not cause a significant impact on the US unemployment rate.  Doesn’t that suggest that those housing construction workers weren’t able to find jobs in other forms of construction, or other activities?

BTW, economic forecasters knew about the drop in housing starts by April 2008.  What sort of unemployment rate did they then predict for 2009?  I’d guess not to high, which indicates economic forecasters probably agree with my view that the housing downturn was not likely to severely impact our unemployment rate.

By the way, commercial RE did eventually decline sharply, as you’d expect when NGDP falls at the fastest rate since 1938.  But that’s not a re-allocation issue, it’s an AD issue.  It does explain, however, why the sharp drop in total construction jobs did lag the sharp drop in residential construction activity.


Tags:

 
 
 

50 Responses to “Bob Murphy wrongly assumes I won’t peek”

  1. Gravatar of Anton Tonev Anton Tonev
    21. January 2011 at 14:20

    excellent paper on this topic by Atif Mian and Amir Sufi (FRBSF) is available here, http://www.frbsf.org/publications/economics/letter/2011/el2011-02.html. The paper looks in general at household debt levels effect on residential spending, durable consumption (both leading indicators in previous recoveries) as well as unemployment. All kind of are interrelated – the areas of US with the highest household debt levels had also the highest unemployment levels and exhibited the highest drop in residential spending as a result of much larger house price declines.

  2. Gravatar of Mark A. Sadowski Mark A. Sadowski
    21. January 2011 at 14:22

    Scott,
    Usually when people post data that undermines their counterarguments it’s because they don’t really understand your original point. On the other hand, data free ideologues often get tripped up by their own data due to lack of experience.

    You wrote:
    “BTW, economic forecasters knew about the drop in housing starts by April 2008. What sort of unemployment rate did they then predict for 2009? I’d guess not to high, which indicates economic forecasters probably agree with my view that the housing downturn was not likely to severely impact our unemployment rate.”

    That’s very easy to check. According to the 2008 Q2 Survey of Professional Forecasters (May 13, 2008) the median projected unemployment rate for 2009 Q2 was 5.5%. Moreover there was this interesting little tidbit:

    “Forecasters See Home Prices Declining Over the Next Two Years and Rebounding in 2010

    In a special question in this survey, the Federal Reserve Bank of Philadelphia asked its panelists for their projections for home prices, as measured by the S&P/Case-Shiller home price index, the OFHEO house price index, and the OFHEO purchase only index…..”

    http://www.philadelphiafed.org/research-and-data/real-time-center/survey-of-professional-forecasters/2008/spfq208.pdf?CFID=16794873&CFTOKEN=96917254&jsessionid=ac30385e581346537c13546a2db205915765

    All eyes were on the housing fiasco in the spring of 2008 and yet forecasters did not expect it to significantly impact unemployment.

  3. Gravatar of Greg Ransom Greg Ransom
    21. January 2011 at 14:56

    Here are the contractors that were in my new house _AFTER_ completion and move-in:

    – roofers

    – painters

    – cabinet guys

    – landscape guys

    – shutter makers

    – pipe fitters

    – drywall guys

    And here are others which worked at my neighbors for more the 2 years AFTER move-in:

    – counter-top guys

    – deck builders

    – floor guys

    – brickwork guys

    And I’ve forgotten dozens of others.

  4. Gravatar of Greg Ransom Greg Ransom
    21. January 2011 at 14:59

    With cross-hairs on Ladera Ranch, CA — AKA Zombieland:

    http://money.cnn.com/2010/12/07/real_estate/ladera_ranch_foreclosure.moneymag/index.htm

    Anton wrote:

    “the areas of US with the highest household debt levels had also the highest unemployment levels and exhibited the highest drop in residential spending as a result of much larger house price declines.”

  5. Gravatar of Greg Ransom Greg Ransom
    21. January 2011 at 15:03

    Check the Orange County Register for unemployment numbers by sector in this ground zero of the housing boom and bust — what you find is a crash in employment in housing finance, real estate, and industry after industry tied to the housing boom (and the spending associated with 2nd mortgages — which came to make up over half of all mortgage finance deals).

    I know these people. They were and are my neighbors. I know a half dozen unemployed real estate agents just on my street.

  6. Gravatar of Greg Ransom Greg Ransom
    21. January 2011 at 15:07

    This isn’t an empirical argument. It isn’t even good economics.

    It’s argument by assertion and fiat.

    When need a breakdown by geography and by sector — in other words, we need people thinking economically and doing field research like scientists in the real sciences, e.g. botany or zoology.

    Scott wrote:

    “commercial RE did eventually decline sharply, as you’d expect when NGDP falls at the fastest rate since 1938. But that’s not a re-allocation issue, it’s an AD issue.”

  7. Gravatar of Dirk Dirk
    21. January 2011 at 15:10

    Despite what data they may use, how do we know that economic forecasters do a better job than a coin flip? Is there evidence? I wonder if their models are merely curve fitted in the same way a lot of stock traders are — so they retrospectively seem to work but actually have no predictive value.

  8. Gravatar of Greg Ransom Greg Ransom
    21. January 2011 at 15:13

    Scott, no one ever asserted that the market discoordination of the boom / bust involve only construction workers.

    Why pretend?

  9. Gravatar of Doc Merlin Doc Merlin
    21. January 2011 at 15:26

    Scott question for you:

    1) If the government pushes AD growth beyond the long term equilibrium growth path will it eventually fall back to the long term equilibrium growth path.

    2) If this doesn’t happen, then why don’t we monetarily stimulate all the time? Don’t say inflation, because inflation is mostly just a problem if its large enough to damage RGDP growth and if money supply is expanding that fast then we aren’t actually boosting AD anymore we are hampering it.

    3) If this (1) does happen, then why do you seem to be against the idea that too much monetary expansion during booms causes the following recession to be more severe?

  10. Gravatar of scott sumner scott sumner
    21. January 2011 at 16:46

    Thanks for the link Anton.

    Mark, Thanks for confirming my assumption.

    Greg, You are still missing the point. If housing completions fell sharply, then clearly much less labor was needed. Yes, there are things to be done after “completion,” but 90% of value added is done by completion.

    I don’t dispute that the recession is worse in areas where the housing crash was more severe, it was also worse where the auto crash was more severe, and the crash in other particular areas with industries like RVs and pleasure boats. I am focusing on the national economy.

    You are wrong about commercial RE, I pointed out it fell precisely when NGDP collapsed, not when housing collapsed. The structural people have no explanation for that timing.

    Dirk, I agree about forecasters, my point was that the knowledge of the housing collapse does not allow one to forecast unemployment.

    Greg, You said;

    “Scott, no one ever asserted that the market discoordination of the boom / bust involve only construction workers.”

    You are putting words into my mouth that I never said. I can quote lots of people who said the housing collapse caused unemployment to soar. That’s what I am responding to, nothing more.

    Doc Merlin, You said;

    “3) If this (1) does happen, then why do you seem to be against the idea that too much monetary expansion during booms causes the following recession to be more severe?”

    Why do you say I am against this idea? It should be obvious I am in favor. I often point to how the Great Inflation of the 1970s led to the big recession of 1982. So I agree.

  11. Gravatar of Morgan Warstler Morgan Warstler
    21. January 2011 at 17:35

    “Why do you say I am against this idea?”

    Because getting you to spend EVEN 1/10th of the time talking about the need to piss on booms, that you spend talking about printing money is about as impossible as getting you to abuse DeKrugman mercilessly for being wrong about Fiscal.

  12. Gravatar of Bob Murphy Bob Murphy
    21. January 2011 at 18:55

    Scott,

    What did you think I meant when I said this:

    That fact alone buys us another few months,

    I know Austrians aren’t good at math, but did you think I believed that from January 2006 to April 2008, only a few months had passed?

  13. Gravatar of Bob Murphy Bob Murphy
    21. January 2011 at 18:59

    Mark Sadowski wrote:

    Usually when people post data that undermines their counterarguments it’s because they don’t really understand your original point. On the other hand, data free ideologues often get tripped up by their own data due to lack of experience.

    I’m not certain, but it seems you are talking about me here, Mark. If so, it’s ironic, since Scott completely misunderstood my point.

    I am really amazed that nobody in the comments has yet wondered why I said “buys us another few months.” Please try to understand what I was arguing, before rushing to the conclusion that I am an idiot and/or ideological hack.

  14. Gravatar of Markus Markus
    21. January 2011 at 20:54

    Scott,

    I’d like to continue the island story from the other post if you don’t mind.
    You said:
    “Markus, It’s a given you can have people build the wrong stuff. But that doesn’t explain unemployment, which is what I’m trying to explain. Your island doesn’t have unemployment.”

    The answer is simply and you said it yourself many times: Sticky wages. Nobody argues with that. If the people on my island have fixed nominal wages they will get unemployment in their situation. Since the farms ran out of food reserves they don’t have enough stuff to sell to pay for the wages. If the wages are sticky the only solution is to lay off people, even though farm products are the most needed in their economy right now. But the cause wasn’t monetary but wrongly allocated resources combined with legal failure (just as Paul Krugman’s baby-sitting piece where you also had a price system that couldn’t adjust).
    Now if you said “Okay we have that sticky wages problem, and misallocations combined with sticky wages lead to massunemployment. But there is nothing we can do about the wages because trade unions are too powerfull and nominal wage cuts would cause civil unrest. So our second best choice is monetary stimulus.” I would be inclined to agree with you. But that’s not what you are doing. Just because you want to propose a monetary solution doesn’t mean the cause of the problem must be monetary. But aparently you think that.
    Just say that misallocations combined with sticky wages and shored off asset prices cause economic downturns. But since we can’t do much about the wages monetary stimulus is our only chance.
    That would be a much more convincing story than your great NGPD tautology “NGPD fell because monetary policy was too tight. How do I know it was too tight? Because NGDP fell.” and you wouldn’t have to argue crazy things like misallocations aren’t even possible, we didn’t build enough houses, and markets are always efficient.

  15. Gravatar of Bob Murphy Bob Murphy
    21. January 2011 at 20:57

    For anyone who’s interested, here’s my response to Scott’s post.

  16. Gravatar of Morgan Warstler Morgan Warstler
    21. January 2011 at 21:53

    “Now if you said “Okay we have that sticky wages problem, and misallocations combined with sticky wages lead to massunemployment. But there is nothing we can do about the wages because trade unions are too powerfull and nominal wage cuts would cause civil unrest. So our second best choice is monetary stimulus.” I would be inclined to agree with you. But that’s not what you are doing. Just because you want to propose a monetary solution doesn’t mean the cause of the problem must be monetary. But aparently you think that.”

    Which is exactly why, if you want a monetary solution you MUST join the team fighting and stomp sticky wages and public employee unions with every damn breathe in your body.

    Progressives KNOW employees aren’t worth what they are paid, but they don’t care, they will lose power if wages go down nominally.

    You KNOW they aren’t worth what they are paid, so you want to simply inflate their earnings away.

    Conservatives KNOW they aren’t worth what they are paid and simply wish to be honest.

    You are in the middle Scott, and while you can win over conservatives, you will not be allowed to make policy unless you help kick liberals out of power.

    See: Uncle Milty.

  17. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 04:51

    Bob,
    You wrote:
    “‘That fact alone buys us another few months’,

    I know Austrians aren’t good at math, but did you think I believed that from January 2006 to April 2008, only a few months had passed?”

    The problem is that those few months go the wrong way.

    Let’s review the numbers:

    January 2006: starts 2,303,000 completions 2,058,000 average 2,180,000

    April 2008: starts 1,008,000 completions 1,014,000 average 1,011,000

    October 2009 starts 527,000 completions 745,000 average 636,000

    Scott’s actually doing you a favor by averaging. The completion data is more damning than the start data.

    The decline in starts between January 2006 and October 2009 is 1776 thousand. The decline in starts from January 2006 to April 2008 is 1295 thousand or about 73% of the overall decline.

    The decline in completions between January 2006 and October 2009 is 1313 thousand. The decline from January 2006 to April 2008 is 1044 thousand or about 80% of the overall decline.

    So the issue is not that completions data moves it only a few months, it’s that it moves it in the wrong direction.

    My advice is give it up before you dig yourself in deeper.

  18. Gravatar of Todd Todd
    22. January 2011 at 05:52

    I’m not a construction industry expert, so someone with more knowledge about how things actually panned out can correct me, but isn’t it possible to use these numbers to support Murphy’s theory? If completion falls faster than starts, couldn’t that mean that fewer workers have left the industry than we might have expected? Maybe houses that should’be taken 6 months to build started taking 9 months.

  19. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 08:44

    Todd,
    I don’t think you need to be an expert to answer that question. It’s just mathematics. If completions fell faster than starts that actually implies the opposite, that the time to complete a house probably decreased. Just imagine a process that grinds to an absolute halt (an extreme case) and you’ll see what I mean.

    And actually it’s intuitive that this is probably exactly what happened. As housing construction wound down that freed up more labor to complete houses faster.

  20. Gravatar of Todd Todd
    22. January 2011 at 09:31

    Mark, that makes sense. The reason I qualified my conjecture, is because I was uncertain as to what counts as a “completion”. If a home is abandoned mid-stream, how does that appear in the data?

  21. Gravatar of Greg Ransom Greg Ransom
    22. January 2011 at 09:50

    The cost of labor went way up during the housing boom.

    Quality dropped.

    Lots of things changed. None of that is of interest to macroeconomists, who think in terms of aggregates complete untethered from changing markets, scarcities, and market processes.

    This is the ultimate legacy of Keynes and the econometricians …

  22. Gravatar of scott sumner scott sumner
    22. January 2011 at 10:06

    Bob. I thought you meant my argument would be weaker if using completions rather than starts, as completions tend to occur a few months after starts. During a slowdown in construction, completions are likely to be higher that starts. I think that’s what the average reader would assume. I never suggested this was your only argument against my position, or even the most important. Indeed I specifically included the very next line “But more important” to indicate that you didn’t think it was your most important point. But surely it counted for something, or else why include it?

    markus, You said;

    “Now if you said “Okay we have that sticky wages problem, and misallocations combined with sticky wages lead to massunemployment. But there is nothing we can do about the wages because trade unions are too powerfull and nominal wage cuts would cause civil unrest. So our second best choice is monetary stimulus.” I would be inclined to agree with you. But that’s not what you are doing. Just because you want to propose a monetary solution doesn’t mean the cause of the problem must be monetary. But apparently you think that.”

    I’m only suggesting a monetary solution in the sense that I am opposed to having monetary contractions causing a fall in NGDP. If monetary policy keeps NGDP rising at a steady 5% rate, I don’t consider that a “solution” I consider it a stable policy. People talk as if my attitude is “oops I see we have a recession, better pull out the monetary solution.” That’s not my policy. If NGDP growth was under 5% I’d be complaining even if we had 5% unemployment. And if unemployment was 20% and we had 6% NGDP growth, I’d say tighten policy.

    My view is that 5% NGDP growth is what the Fed should be doing whether or not there is a housing crisis. If they do that, we will tend to have only mild recessions, not severe recessions. The mild recessions will be caused by the sort of mis-allocation that occurs all the time in market economies, and non-market economies.

    You said;

    “That would be a much more convincing story than your great NGDP tautology “NGDP fell because monetary policy was too tight. How do I know it was too tight? Because NGDP fell.”

    I notice nobody criticizes Friedman for the tautology that money is tight because M2 fell, and M2 fell because money is tight. And I notice no one criticizes Keynesians for the tautology that money is tight because interest rates are high, and interest rates are high because money is tight.

    I’m trying to explain the fall in RGDP, and it is certainly not a tautology that a fall in NGDP causes a fall in RGDP.

    Mark, Yes, that’s a good point.

    Todd, The problem is that completions fell more slowly than starts.

  23. Gravatar of scott sumner scott sumner
    22. January 2011 at 10:07

    Greg, The reason the quality of housing labor is of no interest to me is that I am not trying to explain changes in the quality of labor, and it has little bearing on my recent posts.

  24. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 10:10

    Todd,
    Yes, that is a very good point. I believe such projects end up in limbo (started but not completed). It would be very interesting to see data on abandoned construction projects right now, but, unfortunately, I suspect that such data is not easily produced.

  25. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 10:18

    Scott,
    You wrote:
    “Todd, The problem is that completions fell more slowly than starts.”

    I think what you mean is that they didn’t fall as far. As you agreed with me above I think you acknowledge the adjustment to the current level occurred faster and that this supports your argument with Bob.

  26. Gravatar of Jon Jon
    22. January 2011 at 11:00

    Scott remarks:

    Bob. I thought you meant my argument would be weaker if using completions rather than starts, as completions tend to occur a few months after starts. During a slowdown in construction, completions are likely to be higher that starts. I think that’s what the average reader would assume. I never suggested this was your only argument against my position, or even the most important. Indeed I specifically included the very next line “But more important” to indicate that you didn’t think it was your most important point. But surely it counted for something, or else why include it?

    Because its rhetorical. He is saying you’re being sloppy by not only picking a misleading indicator but by goosing it up by further weakening the correlation by way of picking something which is surely leading.

    You owe Bob an “update”, on this post since you imputed a claim of dishonesty that doesn’t hold up.

    I’d like you to respond directly to his graph of employment in the construction industry vs. unemployment overall. Keep in mind that under the Misesian narrative, its a tightening of credit that eventually induces the liquidation.

  27. Gravatar of Dan Dan
    22. January 2011 at 12:21

    Scott,

    How does your theory explain the three graphs that were in Bob’s post?

  28. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 17:15

    Dan,
    The real question is how Bob’s three graphs relate to Scott’s theory. If you can explain that, we might be halfway there.

  29. Gravatar of Dan Dan
    22. January 2011 at 18:00

    Mark,
    That is the question I just asked Scott.

    If you are asking why Bob’s graphs are relevant to Scott’s attack on Kling well I couldn’t spell it out any better than Bob’s article did. Besides if Scott thought Murphy put up graphs that were irrelevant or wrong he could have blew the article up. Instead he took one point Bob made and expanded on that further as if that were a refutation. Bob tried to explain that construction jobs are more than just housing starts. He used housing completions as one example showing and saying that this data bought a couple months. He also said that construction in commercial real estate, building decks, etc helped keep construction employment from falling. Then showed three graphs that every Misesian would expect to see from a boom and bust. Sumner comes in and takes the housing completion data, stretches it out over 2 years instead of the few months Murphy relayed, talks about construction moving to commercial real estate as Murphy did, and completely ignores everything else in the article. The only thing I can come up with is that Scott doesn’t have an answer for the graphs Murphy showed.

  30. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 18:08

    Dan wrote:
    “If you are asking why Bob’s graphs are relevant to Scott’s attack on Kling well I couldn’t spell it out any better than Bob’s article did.”

    In other words your explanation as to why those three graphs were relevant would have been as weak as his was.

    Still waiting for a real answer. Good luck with responses like that in persuading anyone.

  31. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 18:17

    Dan wrote:
    “Then showed three graphs that every Misesian would expect to see from a boom and bust.”

    I’m not Misean. The vast majority of the population are not Misean. If you want more people to be Misean then you have to try and make a persuasive argument. Go to it!

  32. Gravatar of Morgan Warstler Morgan Warstler
    22. January 2011 at 19:12

    “Still waiting for a real answer. Good luck with responses like that in persuading anyone.”

    Yeah, it’s 100% compelling – my entire family are vertically integrated into housing – from development to home homebuilding, to interior design to legal stuff… my father has a number of crews.

    By my read, the best answer was in Scott’s first post from OGT:

    “Because of the changes in household size, the U.S. needed far more additional housing units in the ’70s and ’80s than in the ’90s and ’00s. If we could normalize the housing start chart by household formation, we would see that the last decade was indeed extraordinary!”

    http://www.calculatedriskblog.com/2010/04/housing-impact-of-changes-in-household.html

    Scott responded by saying, “that assumes away the problem,” and prattled on a bit about 20 somethings living at home.

    And THAT’s the whole of it:

    1. The trend line in Scott’s first assertion was wrong… trust Calculated Risk here.
    2. So the growth from 2000-2008 was too high – meaning Scott was assuming away the problem by pretending we need more houses than we do.
    3. This is not to say, that IF we had built less houses as we tightened money up a bit starting in 2004, that we’d not have less unemployment now and more 20 somethings out buying houses from a pool about the same size as NOW, we’d just have been building 1.2M per month from 2004 onward, instead of 1.8M then, and 6000K now.

    So if thats the reality, then Scott trying to play gotcha on unemployment numbers pegged hard-core to housing starts makes less sense…. because in my experience home builders started taking remodeling work on existing homes, they started working on blocks of $125K townhouses that still were cheap under FHA 3% down – trying to woo those kids out of their parents house… overall, the industry wasn’t ready to let go.

    One more thing, looking at the Calculated Risk Numbers here:

    http://calculatedriskimages.blogspot.com/2010/04/housing-starts-march-2010.html

    What looks REALLY odd, is the way that during the last housing boom – the make up of single homes to multi-unit is quite different.

    During those past housing booms – like 30-50% of the starts are in multi-units, you could almost say in past boom they really just built a lot of multi-tenant.

    And then in 1994 onward, we don’t see that – suddenly for the first time single unit homes are the bulk, and then when the 2002 boom happens – they stay the bulk.

  33. Gravatar of Bob Murphy Bob Murphy
    22. January 2011 at 21:01

    Hi everyone,

    I see this is still a lively topic. Mark and I went toe-to-toe in the comments at my post responding to Scott, but let me give a general reaction here, since I like away games as much as playing at home:

    (1) Scott showed that housing starts dramatically fell from January 2006 through April 2008. However, during the same period, the national unemployment rate hardly moved. Scott said (possibly paraphrasing), “Looks like those construction workers found something else to do…” One would have thought that construction employment surged into 2006, then fell sharply, even while national unemployment didn’t move. So clearly the Recalculation Story was in jeopardy.

    (2) I pointed out that, if this is what Scott and others inferred, they would be wrong. Construction employment peaked and held steady all through the period when the national unemployment rate held steady, and then when construction employment tanked, the national unemployment rate zoomed upward. So the basic Klingian story was back in play–it looked like we could explain the national unemployment problem, by the disgorging of construction workers out of that sector.

    (3) OK, but there was an obvious problem: How to relate this to the housing boom? Arnold Kling and others (including me) who say the recession was “real” and not “nominal” often say to the layperson things like, “Too many workers and lumber went into housing. Workers lost their jobs in Nevada and it takes time for them to go somewhere else.” So the problem for us as Recalculation/Austrian theorists is, can we reconcile our original story with the new facts? I.e. we’re on solid ground with respect to total construction employment, but we’ve been talking as if all these workers were building houses. And yet, as Scott showed, housing starts started falling fast after January 2006, whereas we know construction employment held steady for a long time after that. What gives?

    (4) Well, one thing is that houses aren’t made in a day. So when we switch to housing completions, we gain “a few” months. (I think it’s technically only “a couple of” months, but since it peaks in March I thought “three months after the start of 2006.”)

    (5) Clearly two or three months isn’t going to cut it, because we need to explain the steady construction employment through April 2008 at least. So that’s why I went into the discussion about things that would be related to a massive surge in new home construction, such as adding decks and building shopping malls near the new developments.

    (6) When writing up this argument, I made almost an offhand remark about the distinction between housing starts and completions, and then linked to the data. I said housing completions peaked “a few months” after housing starts. That is undeniably true (if you give me an out on “couple” versus “few”).

    (7) In light of the above, I am still surprised that Scott thought I was hiding something in the link. Housing completions did exactly what I said they did. That data doesn’t “bolster” Scott’s case, because I’ve already shown that housing activity by itself didn’t mesh with overall construction employment.

    (8) Finally: If it had turned out that all the people who were building houses in 2006 were picking vegetables and writing novels in 2007, then Scott would be totally vindicated: Rising NGDP solves all problems, and we don’t need to worry about labor skills matching up with consumer demands. But since Scott himself agreed that the former housebuilders moved into other construction projects–which may very well have been corollaries of the previous housing construction–I think Kling still has a leg to stand on. (Not surprisingly, so does Kling.)

  34. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 21:32

    Bob you wrote:
    “But since Scott himself agreed that the former housebuilders moved into other construction projects–which may very well have been corollaries of the previous housing construction–I think Kling still has a leg to stand on. (Not surprisingly, so does Kling.)”

    Let me quote again for emphasis:

    “WHICH MAY VERY WELL HAVE BEEN COROLLARIES OF THE PREVIOUS HOUSING CONSTRUCTION”

    Are you sure you want to make this the crux of your incredibly weak argument?

    If so prove it!

  35. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. January 2011 at 21:54

    By the way Bob I routinely ignore Morgan so don’t take any notice of it.

  36. Gravatar of Stefan Karlsson Stefan Karlsson
    22. January 2011 at 23:49

    I have explained here what both Sumner and Murphy missed in this debate.

  37. Gravatar of Silas Barta Silas Barta
    23. January 2011 at 07:56

    @Mark_A._Sadowski:

    I thought the weak argument would be trying to argue that it doesn’t matter if worker skills match what consumers want as long as people have enough dollars?

  38. Gravatar of Morgan Warstler Morgan Warstler
    23. January 2011 at 08:29

    Mark, I’m just hear to make sure you get pulled into Scott’s orbit (which is is very much my own).

    OK, back to the housing starts vs. single units.

    http://calculatedriskimages.blogspot.com/2010/04/housing-starts-march-2010.html

    What’s interesting is that the building of single units peaked previously in 1978-79 at 1.5M… so if we had started to freak out about single starts getting too hot, based on historical data, Jan 02 would have been the time to start paying attention.

    What’s more look at the little blip 80-82, that’s probably the nicest little proof for the basic rule we see until about the same time Jan 02…

    Real boom growth comes in multi-tenant. The hotter it gets a bigger the percentage of the growth is multi-tenant.

    Did this have something to to with FHA / F&F?

    Maybe during previous booms the financing went to developers who took the risks themselves building apartment / condos?

  39. Gravatar of The Window Washer The Window Washer
    23. January 2011 at 08:53

    Something everyone in construction noticed in 07-08 was the mexicans went home. I would say that the workers on jobs went from 50 to 20% hispanic. Thats in high end multi million dollar single family. I dont’ know about tract home.
    This was something everone was talking about because the jobs numbers wasn’t showing anything. Once the “construction job losses” started hitting the news momentum was already moving very fast in the industry.
    The route to layoffs in the trades is unskilled illegals, unskilled/problem legals, skilled illegals, experienced skilled legals then busness shutdown right after the experienced workers go.

  40. Gravatar of Mike Sandifer Mike Sandifer
    23. January 2011 at 09:51

    Morgan,

    Again, you’re focusing on something which may as well be a non-issue. The costs of government wages and benefits don’t matter much if the Feds do their jobs and stimulate the economy.

  41. Gravatar of Dan Dan
    23. January 2011 at 12:04

    Mark,

    Well I can see why you wouldn’t get those graphs and how they back up ABCT since your man Sumner chose to leave you in the dark by not even responding to them. If you were hoping to learn ABCT by reading a post or blog I’m sorry but it isn’t that easy. Go to Mises.org and do as much research as you want.

  42. Gravatar of David Beckworth David Beckworth
    23. January 2011 at 14:54

    All:

    Here is a post
    I just did on this issue. No surprise where I come out, but I do present more data

  43. Gravatar of Mark A. Sadowski Mark A. Sadowski
    23. January 2011 at 18:06

    Silas Barta,
    You wrote:
    “I thought the weak argument would be trying to argue that it doesn’t matter if worker skills match what consumers want as long as people have enough dollars?”

    No, the weak argument is stating that there is a sectoral mismatch when the current employment data and the compensation data suggest otherwise. Sectors accounting for over 85% of employment have suffered declines. When you (or Bob) bother trying explaining how “recalculation” or vulgar Austrianism is consistent with this fact it will be a red letter day.

    No one is arguing that “recalculation” can’t ever happen, only that it doesn’t match the current data.

  44. Gravatar of scott sumner scott sumner
    23. January 2011 at 19:30

    Mark, Thanks.

    Jon, I don’t understand your argument. I think any reader of Bob’s post would assume I picked starts to make my case look better. But it doesn’t, it makes my case look worse. Using Bob’s data makes it look better. I did an update pointing that out. It’s this post. And now you want another update pointing out . . . what?

    Dan, Bob shows that as construction employment falls sharply, total unemployment rises. That’s perfectly consistent with my model, which claims the recession caused a huge drop in non-residential construction.

    You said;

    “He used housing completions as one example showing and saying that this data bought a couple months.”

    But as my post shows that’s wrong, it buys zero months. Indeed makes my argument stronger.

    Morgan, You still don’t get it; the stronger the economy the more households. Do you think anyone in 2007 predicted that in the next few years 550,000 units a a year would meet all new household formation? I bet not one demographer made that forecast. Indeed it’s absurd if you think about it. Demographic trends change slowly, not all at once.

    Bob, You said;

    “(2) I pointed out that, if this is what Scott and others inferred, they would be wrong. Construction employment peaked and held steady all through the period when the national unemployment rate held steady, and then when construction employment tanked, the national unemployment rate zoomed upward. So the basic Klingian story was back in play–it looked like we could explain the national unemployment problem, by the disgorging of construction workers out of that sector.”

    What you showed has no relevance to my argument. I never argued total construction employment fell, I argued residential construction employment fell. Elsewhere Greg Ransom has lots of data showing unemployment rose more sharply in counties where the housing bust was more severe, so residential construction firms were shedding jobs in 2007–it just wasn’t have much effect on the unemployment rate. Not zero effect, but a modest effect. The big rise in unemployment occurred in late 2008, after the housing bust was mostly over. In mid-2008 forecasters did not forecast high unemployment for 2009, even though they had the data on the huge plunge in housing starts. A housing slump, by itself, is not enough to cause a major rise in unemployment. I fully agree that when the fall in NGDP caused jobs to tank in commercial RE, things got substantially worse. That’s what your graphs show.

    You said;

    “4) Well, one thing is that houses aren’t made in a day. So when we switch to housing completions, we gain “a few” months. (I think it’s technically only “a couple of” months, but since it peaks in March I thought “three months after the start of 2006.”)”

    This is completely wrong. As my post shows you don’t gain any months, completions makes my argument stronger.

    You said:

    “In light of the above, I am still surprised that Scott thought I was hiding something in the link.”

    Well the hiding part was a joke. Like when Kling said I was fighting dirty.

    Mark, yes, I fail to see how commercial RE, which peaked several years after housing starts, was a corollary of the housing boom. Shopping malls sell to people, not to houses. And even if they sold to people, the timing is all wrong. And commercial RE prices peaked in late 2008, evidence the market didn’t know a crash was coming. If it was caused by less housing, they would have known.

    Stefan, Can you summarize over here?

    The window washer, That may be true, but I see it as supporting my point that the housing decline didn’t have much affect on unemployment between 2006-08.

    Thanks David.

    Mark, Yes, almost all sectors have seen big declines, as you’d expect when AD falls sharply.

  45. Gravatar of scott sumner scott sumner
    23. January 2011 at 19:40

    Stefan, I think your data on the boom in commercial RE during 2006-08, strongly supports my view that residential construction employment probably fell substantially during 2006-08.

    Because you started your peak in 2005:4, rather than Jan. 2006, the decline looks smaller. That’s partly because the peak was lower in your data. It’s not because I use housing starts, using completions makes my argument even stronger.

    David, Those are great graphs, really shows how the Great Recession was all about NGDP, not housing construction.

  46. Gravatar of Mark A. Sadowski Mark A. Sadowski
    23. January 2011 at 21:27

    Morgan wrote:
    “Mark, I’m just hear to make sure you get pulled into Scott’s orbit (which is is very much my own).”

    You don’t have to worry about my orbit Morgan.

    But I often worry about the orbit of people like you. That’s not because of your intentions (which I believe are pure). It’s for your effects (which I am sure are unintended).

  47. Gravatar of Pop Quiz on Economists! Pop Quiz on Economists!
    24. January 2011 at 04:02

    [...] in case you have had the flu or something, I’ll give you, let’s say, 3 minutes to skim this post and get up to speed with where we’ve taken the class this past [...]

  48. Gravatar of Gregor Bush Gregor Bush
    24. January 2011 at 06:22

    Scott,
    You can break down construction employment into residential and non-residential sub-sectors. At its peak, residential construction accounted for a grand total of 0.75% of US employment. By April 2008 it was 0.62%.
    Employees working in “real estate services” accounted for 1.1% of emplyment at the peak and this number has stayed roughly constant through the crisis and recession.

    In terms of absolute numbers, the combined decline in residential construction and real estate services from the peak in March of 2006 to today is 579,000 (and that of course includes endogenous impact of the financial crisis and the recession itself). So how in the world to Arnold and Bob square this with a peak-to-trough employment decline of more than 7 million in an economy with trend labour force growth of at least 1.5 million per year?

    The declines in manufacturing, retail and professional and business services employment from August of 2008 to Decemeber of 2009 were 1.8 million, 878K and 1.1 million, respectively. residential constrcution employment over this period fell only 220K. I don’t know how anyone could possibly believe that “recalculation” has been the primary driver of employment delines after looking at the data.

  49. Gravatar of scott sumner scott sumner
    24. January 2011 at 09:34

    Gregor, Your numbers include only construction workers, not trades contractors who who also help build houses. See my new post. It’s more like 2% of workers.

  50. Gravatar of TheMoneyIllusion » Why hasn’t the housing recovery helped the economic recovery? TheMoneyIllusion » Why hasn’t the housing recovery helped the economic recovery?
    21. January 2013 at 07:10

    [...] Most people are convinced that the housing crash contributed to the recession, even though the facts suggest otherwise. [...]

Leave a Reply