How Much Did the Stimulus Stimulate?

Brad DeLong takes note of a recent paper evaluating the results of the ARRA, the stimulus package. Quoting from the paper:

A cross state analysis suggests that one additional job was created by each $170,000 in stimulus spending. Time series analysis at the state level suggests a smaller response with a per job cost of about $400,000. These results imply Keynesian multipliers between 0.5 and 1.0, somewhat lower than those assumed by the administration.

to which Dr. DeLong responds that

All of these estimates are what we call “imprecisely estimated”: whatever your prior beliefs were, they should not move very much.

My “prior belief” is that a well-crafted stimulus package applied in a timely fashion might well have produced a higher multiplier and, if we were a compact country with a small, homogeneous population ruled by a totalitarian regime with complete power, we could have produced one. Unfortunately, as a large, sprawling country with diverse competing interests and serious political differences we’re incapable of producing a well-crafted stimulus package and applying it in a timely fashion. Consequently, we’ll see smaller multipliers here than theory might dictate.

On a related note I see that China has hiked interest rates again:

LONDON (MarketWatch) — China’s central bank announced Tuesday that it would raise its key interest rates for the third time since October in an effort to cool rising inflation pressures.

The People’s Bank of China said in a statement that it will raise its one-year yuan lending rate to 6.06% from 5.81% effective Wednesday, while boosting the one-year yuan deposit rate to 3% from 2.75%. Rates were previously hiked in October and December.

suggesting that China’s stimulus package might have produced some hard to tame side effects.

18 comments… add one
  • steve

    It appears that the block grants to the states were the weak part of the stimulus. Was that predictable? Would not one ordinarily expect that spending at a more local level would be done more wisely?

    Steve

  • PD Shaw

    As I understand it, a multiplier of less than 1 means contraction in the private sector has occurred as a result of government policy. So, I don’t see this as a “nothing to see here, move along” conclusion.

    I’m not sure that a less than one multiplier is an unvarnished negative, it’s just that some sectors have been helped other’s hurt. DeLong’s charts indicates where.

  • PD Shaw

    steve, I think the problem was that the theory behind a succesful stimulus is that it increases total government spending, so the states and local government cutting back on spending needs to be offset by more federal spending. So, I believe the extent the federal grants merely offset state and local government cuts, the multiplier would tends towards zero.

  • Drew

    “My “prior belief” is that a well-crafted stimulus package applied in a timely fashion might well have produced a higher multiplier and, if we were a compact country with a small, homogeneous population ruled by a totalitarian regime with complete power, we could have produced one. ”

    I’m sure jp is working on a remix of the old “I’m turning Japanese, I think I’m turning Japanese, I really think so” song to “I’m turning Danish……. ”

    But seriously folks. As Dave knows, because I sent him the link at the time the stimulus was being debated, at a UofC economics forum my old prof Kevin Murphy put forth a model showing how the multiplier could be anemic, or negative. It was a tad tongue in cheek, but of course the lynchpin was the effectiveness of spending variable. And we all know how that turned out. And now we have helicopter spending to prop things up……..super.

    steve – given that the local issue is out of control pensions, salaries and health care costs…………….No.

    One of the things I love about my business is that it gets you front and center of issues in many different businesses and industries. You learn first hand what’s really going on; not what’s written in some book, on the cable channels or said by some commentator. We actually sold a portfolio company last year because we believed that municipal spending was going to be a huge problem – politician’s rhetoric aside; and the pols have played true to form: spending on public sector workers compensation, er, reliable voters issues, before hard assets. The company had a heavy reliance on school spending.

    So much for the supposed focus on education.

    “I’d like to help you son, but you’re too young to vote.” “Sometimes I wonder, what I’m a-gonna to do ’cause there ain’t no cure for the summertime blues……”

  • Considering that DeLong is not a Bayesian his prattle about prior beliefs is a bit…amusing (in an intellectually vacuous manner).

    As I understand it, a multiplier of less than 1 means contraction in the private sector has occurred as a result of government policy. So, I don’t see this as a “nothing to see here, move along” conclusion.

    Well that is why DeLong is suddenly seeing the Bayesian light all of a sudden.

  • PD Shaw

    Here is apparently the paper’s explanation of the problem with the state and local payments:

    “Transfers to state and local governments didn’t work, the paper said, because states considered the grants to be ‘temporary,’ and therefore avoided hiring full-time workers. ‘States may have used the money to lower borrowing or limit tax increases.’”

    http://www.nationalpost.com/news/canada/recovery+worth/4243750/story.html#ixzz1DOdnuuXE

    The models used by Romer and Bernstein to get higher multiplier effects assume government spending is increased permanently. If governments, households and businesses assume that the stimulus is a one-time event, they are going to behave differently.

  • steve

    Then the missing part is the number of jobs that would have been lost if states had not received their grants. Would they have been willing to increase taxes to maintain the same number of jobs? Would that have decreased growth?

    Steve

  • Drew

    “Then the missing part is the number of jobs that would have been lost if states had not received their grants. Would they have been willing to increase taxes to maintain the same number of jobs? Would that have decreased growth?”

    Ahaa! Now we’re talkin’. Let me restate the question: SHOULD taxes have been increased to maintan the ame number of jobs?

    And instead of “would that have decreased growth,” how about “would anyone have noticed, and could those resources have been more productively used to create growth in the private sector??”

    Being the contrarian prick I am, I ask anyone: so we’ve been in recession for 3 years. The brunt has been taken by the private sector. The public sector has done relatively quite well, what with their guaranteed jobs, salaries and pensions…………….how’s that workin’ out? Your driver’s licenses comin’ out more quickly? IRS more efficient or friendly? Roads and bridges cheaper? Sales taxes down? Property taxes? Mr Private Sector – 401K? Job?

    Gettin’ tired of workin’ for “the man?” But “the man” is the government?

  • sam

    “The public sector has done relatively quite well, what with their guaranteed jobs, salaries and pensions”

    Evidently the libertarian paradise known as Texas used the federal stimulus money it received to put a tourniquet on its budgetary hemorrhage. To ultimately no avail .

  • steve

    “Being the contrarian prick I am, I ask anyone: so we’ve been in recession for 3 years. The brunt has been taken by the private sector.”

    At first. I assume you have been following unemployment stats and know that jobs in the public sector have been decreasing while we are gaining in the private sector.

    ““would anyone have noticed, and could those resources have been more productively used to create growth in the private sector??”

    Good question. At the state and local level, most employees are teachers or police, including corrections. While many are willing to complain about teachers, few are willing to fire them. Fire enough, and you create a drag on future growth. As to police, at some point there is clearly a drag on the economy if you dont have enough. Which ignores the fact that in a recession, states are facing an acute revenue problem. Firing those teachers and police do nothing to affect revenue. It would help to balance a state budget, but I cannot see how it would affect growth positively in the short term.

    Steve

  • How much of state budget problems are due to pensions? If it is significant all this talk of firing teachers and police is less of the issue in that the problem lies somewhere else. And then the problem is one of will such problems get bigger going forward? If the answer is yes, then the grants to the states was simply kicking the can down the road and letting it get bigger. Classic example of moral hazard.

  • At first. I assume you have been following unemployment stats and know that jobs in the public sector have been decreasing while we are gaining in the private sector.

    While technically true this is misleading. Using the BLS statistics and setting aside the spike due to the Census, total government employment is off by about 1.5% from its previous high. While that has been a reversing of the trend in total private employment total employment is just about 6.6% off of its recent high. So, while you are correct steve about the job gains, Drew has the better of the argument.

    Also there is this article,

    http://www.csmonitor.com/USA/Politics/2010/0701/New-state-budget-plans-burdened-by-recession-pension-costs

    Payroll and pension costs. States have cut about 18,000 jobs (less than 1 percent) since the middle of 2008. Local governments have cut more deeply – 158,000 or 1 percent of jobs. Several states have negotiated with unions to boost the retirement age, but this generally affects newly hired workers. One study conducted last year, by Robert Novy-Marx and Joshua Rauh of the University of Chicago, estimated that state pensions are underfunded by $1 trillion to $3 trillion.

    It is a bit old, but it does show that the pension/health care plans problem are the 600 pound gorillas sitting in each state’s budgeting room. And every politician is a complete chicken shit when it comes to facing the issue.

  • sam

    About that Texas libertarian miracle:

    Perry demands more money from Congress for Texas schools

    Gov. Rick Perry, in his State of the State speech Tuesday, demanded that Congress free up $830 million for Texas schools — prompting a denunciation from Democrats who say he just wants the money to help plug the state’s huge budget hole.

    Last summer, Texas Democrats in Congress secured a provision in federal law that says Texas, to get the federal money, must promise not to cut its own spending on education for three years. Perry says he can’t legally promise that, and on Tuesday, he renewed complaints about congressional meddling, calling out Democratic U.S. Rep. Lloyd Doggett of Austin, who authored the provision.

    “The Doggett amendment … is taking more than $830 million from Texas schoolchildren and teachers,” Perry said.

    The story said the the Gov is flying to DC to have a talk with Eric Cantor. (Good luck.) Texas is sitting on a $9 billion “rainy day” fund, and one wonders what the folks in Austin would consider a rainy day to be.

    For info on the Texas public employee pension problem, see
    Texas public pensions under scrutiny in spite of protections. Texas’s pension burden doesn’t seem to be the giant weight that it is in other places. For example, there is no automatic COLA (last time the pensioners got a bump was in 2001).
    So I’m not sure we can attribute Texas’s problems all that much to public employee pensions.

  • john personna

    To extract economic signal from noise you need “all else being equal.”

    What we got was an imperfect federal stimulus spread over a time when a great deal was changing rapidly. In California we got stimulus funds, but laid off teachers at the same time. We built highways, but abandoned housing tracts.

    So what can we reasonably say? I’m sure fewer people lost their jobs than a no-stimulus (austerity) alternative, but more lost their jobs than better (or bigger) stimulus might have done.

    Remember, my idea for stimulus was straight counter-cyclical spending. That was to only move forward projects planned or necessary in the future, and to do more of them during the slump. That was probably a semi-austere plan, and probably would have necessitated more safety net spending.

    But you know, my plan was just as unlikely as any of yours.

    (We can always strive to be more sensible and Nordic, Drew. That’s a given.)

  • In California we got stimulus funds, but laid off teachers at the same time. We built highways, but abandoned housing tracts.

    This highlights what I think is one of the great misdirections of the discussion of the interrelationships among federal spending, state spending, and the economy. If California is anything like Illinois, it didn’t lay off teachers because it didn’t have the money to retain them. It laid them off because it didn’t have the money to maintain its pay schedules for that number of teacher and given the choices between maintaining its pay schedules and laying off teachers, it elected to lay off teachers.

    Here in Chicago teachers with bachelors in education only (which have been notoriously weak programs for more than a half century) start at $45,000 for a ten month a year job from which they can retire at 75% of pay. Police officers and firefighters with high school only educations make around $80,000 per year and can retire at age 50 at 75% pay.

    Framing the discussion as “they had to lay off teachers” is far too simple.

    Also, as I’ve tried to explain before, “sensible” is not an objective term but a subjective one. What is sensible depends on preferences and circumstances.

  • john personna

    Well, my point was more the confusion in changes more than any “connections.”

    FWIW we were told districts lost revenues.

  • sam,

    I think that is rather dishonest to refer to Texas as a libertarian anything.

  • sam

    Yeah, I should have put ‘libertarian’ in shudder quotes.

Leave a Comment