Wednesday, January 19, 2011

Nominal Spending Then and Now

Is it really that hard for monetary policy restore total current dollar spending to trend?  As an advocate of nominal GDP level targeting, I certainly believe the Fed is capable of such a task and have been making the case for sometime.  Skeptics, however, typically throw out some version of the Fed "pushing on a string" argument as a rebuttal. While appealing, this view ignores the best data point we have on this question: the Great Depression.  As is well known, nominal spending fell in half during this time and slowly recovered during the decade. As is also well known, FDR's monetary easing--devaluing the gold content of dollar and not sterilizing gold inflows--and the nominal expectations it created were key to restoring nominal spending.  The recovery path was not perfect, but eventually total current dollar spending returned to its pre-Great Depression trend as seen in the figure below:

This is a remarkable accomplishment given the dire circumstances of the 1930s.  Note that full employment was restored without the unleashing of hyperinflation or any of the other concerns raised by critics of monetary stimulus.  So why can't the same happen today?  Why is it that instead of having a figure like the one above where nominal spending returns to trend we instead get the following one?


  1. David,
    I'll hazard a guess, at the risk of being dismissed as a lightweight. Look at the M1 Multiplier. It is in negative territory. Would this not indicate that a dollar added to the money supply does not in fact add a full dollar in kind? I suspect that expectations on the part of spenders may also be a factor. Uncertainty leads to holding back. During the Depression, there was pent up demand for goods and services, with no means of purchasing power. Just a thought.

  2. Congrats, you got mentioned by Krugman today, along with Sumner. You guys are making an impact (positive in my view) on monetary public policy discussions.

    Here is a non-monetary idea I just had, but maybe it is worth pondering.

    The USA has open borders--capital, goods, services, and even labor flow freely (despite the Keystone Cops antics of the Border Patrol).

    Okay, let's say we get stimulative demand, monetary in origin.

    Demand increase, economic growth ensues. Now, at some point, you should get inflation. But has that point really been pushed way back?

    Strong demand for labor means Latin America shows up at our doorstep, lots of hard-working people. Or, we can establish call centers in India. Capital crosses the border in an instant, supplying business with money for expansion. Capacity is increased.

    Goods obviously pour in through the huge ports at Los Angeles and elsewhere--strong demand for cars, and in come the cars. Little chance for rent.

    It seems to me this open border policy easily thwarts inflation--one reason why there has been so little in the last 20 years.

    Add to that a national glut of all types of commercial space.

    I worry about inflation about as much as I worry about getting beaten up by Liberace.

  3. IIRC, the stimulus was very small, much less than 1% of GDP, compared to 2% during the Great Depression. ~40% was tax cuts, which have a low multiplier (especially for higher-income people). The stimulus at best countered some of the massive state budget cuts (and wouldn't have countered much of the county/city cuts).

  4. It's funny, but I have a chart of nominal spending going back x years, and it shows me that the decline in nominal spending has fallen back to the trend. Things have returned to normal! You have a chart going back x years, and it shows you that the decline in nominal spending has gone below trend. My friend Joe has a chart that say nominal spending needs to fall even further to get back to trend.

    Live by the chart, die by the chart. If you're going to make arguments and use figures as your bludgeons, you have to explain them far better. Start with... why is your period of measurement correct? Why should we care about your trendline? Cause as it is, we can't say you're not cherry-picking whatever trendline supports your argument.

  5. JP Koning:

    That is a fair criticism. I did a post a few days back on this very issue where I chose my trend based on the 1987-1998 period. I explain there why I chose this benchmark period. Also, my preference is to look at measures of final sales, not nominal GDP. I only used nominal GDP here to make the comparison across periods.

  6. Nanute, Benjamin, and Barry:

    I think the fundamental reason is that there is not a political consensus to do so. This is sort of the point Scott Sumner just made at his blog. So

  7. David, thanks for the link. The conversation between yourself, Harless, and Woolsey reminds me of a couple of technical analysts arguing over what moving average best defines a stock trend. Well, you can always find some moving average to serve as a hypothetical resistance level and thereby justify a stock purchase.

    Where on the St Louis FRED website can I find the data series you prefer? What kind of trendline are you using?

  8. JP Koning:

    Final sales of domestic output is probably the closest thing to textbook aggregate demand while final sales to domestic purchasers is domestic demand. To calculate the non-linear trend I simply fitted a regression with a time trend and a time trend squared (i.e. y = a + b1*t + b2*t^2). Alternatively, I could have taken the log of the series and then estimated a linear trend.

  9. Thanks, will check out. Just call me JP.

  10. Good post very interesting, the graphs were good.