George Selgin on Deflation and Nominal Income Rules

Before beginning this post I should mention that I don’t have Selgin’s most important book on deflation (Less than Zero) in front of me, rather I am reacting to a recent article from The American Conservative that George sent me.  Even so, given all the comments that I have received about the relative merits of inflation and deflation, I thought it would be worth discussing his ideas.

George distinguishes between “good deflation” which is associated with technological progress reducing the cost of production, and “bad deflation,” which is caused by a fall in aggregate demand, or what George and I prefer to call nominal spending.  George favors stabilizing nominal spending in order to stabilize incomes.  Under that sort of policy, technological progress would boost real incomes by reducing prices, not raising nominal incomes.  I am not certain whether George favors stabilizing total nominal income (roughly nominal GDP, but George prefers nominal final sales) or whether he favors stabilizing nominal income per capita, which would mean that total nominal GDP would rise by roughly one percent per year.   I will confine my comments to the latter interpretation, as nothing in my argument would change if the former interpretation was correct.

George also cites examples of both good deflation (1873-95) and bad deflation (1929-33.)  The difference is that nominal GDP did not fall in the former case, but did in the latter case.  He also argues that some of our recent problems with the business cycle are due the Fed adopting a two or three percent inflation target, rather than stabilizing nominal income.  You may have noticed that my preferred target, roughly 5% nominal income growth, straddles these two options.  Like George I favor a nominal income target, and like the Fed I think that 2% average inflation is probably better than 2% average deflation.  There are at least 5 issues that are relevant to this debate, the first two favor my view, the next two favor George’s view, and the last favors either of our views over the more widely accepted inflation targeting concept:

1.  Sticky wages

2.  Liquidity traps

3.  Menu costs

4.  Taxation of capital

5.  Macroeconomic stability

I don’t know if I have ever explained the title of this blog, but when economists use the term ‘money illusion,’ they are usually referring to how the public confuses real and nominal variables.  Thus we will have a pay freeze at Bentley this year, whereas last year we had a 3 or 4% increase in nominal salaries.  Of course in the interim inflation has fallen from the 4-5% range to slightly negative, so in real terms we are actually doing better this year.  Nevertheless, many economists assume that the public confuses real and nominal wages (at least to some extent) and feels worse about the zero percent raise this year than they did with the 3 or 4% raise last year.  Of course when inflation rises the public does catch on at some point, they are not completely ignorant of the distinction between real and nominal variables, but some economists also insist that there is a sort of psychological resistance to nominal wage cuts, even if one’s real wage is rising.

George emphasizes that under his policy of nominal income targeting the aggregate income level need not fall (unlike 1929-33.)  But even apart from aggregate shocks, the economy is continually buffeted by various sectoral disturbances, and these changes are partly accommodated by adjustments in relative wage levels across industries.  If the aggregate nominal wage were stable, then nearly half of all workers might need to take a nominal pay cut during any given year.  This may or may not be a big problem, but I want to emphasize that “money illusion” doesn’t seem to be merely something dreamed up by economists, the distribution of nominal wage changes shows a sharp discontinuity at 0%, which is hard to explain in any model excluding money illusion.

(BTW, although this concept is what most economists mean by the term “money illusion,” my blog title has more than one layer of meaning, and the most important is hinted at in my posts on how economists are misreading the current stance of monetary policy, and the prospects that policy could do more to boost demand.  Indeed, I think that most of what we “know” about monetary policy—such as that policy affects the economy by changing short term interest rates—is an illusion.)

Because the U.S. has recently experienced roughly 5% nominal GDP growth, and because monetary policy affects only prices (not real output) in the long run, we can assume that under George’s proposal inflation would be roughly 4% lower, on average, than the rates experienced in recent decades.  In addition, the Fisher effect tells us that nominal interest rates would also be roughly 4% lower than the actual rates that we have experienced.  That is, nominal interest rates would be very low.

This is not necessarily a big problem, and might (as Milton Friedman argued) even be an advantage for mild deflation.  My concern is that under current operating procedures we simply don’t know how to conduct monetary policy in a near-zero interest rate environment.  I can imagine George responding in several possible ways.  First, our current policy predicament is partly due to the fact that we have not followed a nominal income rule, and second, monetary policy can still work in a zero rate environment if we don’t rely on the short term interest rate as the policy instrument.  I don’t know whether he would make these arguments, but if he did I would find both of them very plausible.  But even so, we are dealing with existing governmental institutions that clearly don’t seem comfortable operating in a zero interest rate environment, so I still worry about a stable nominal income rule, but more for pragmatic reasons rather than theoretical reasons.

On advantage of George’s proposal over mine is that the so-called menu costs would be a bit less.  The difference between 2% deflation and 2% inflation is not significant in this regard, but under Georges’ proposal there would be less need for frequent adjustments in nominal wage rates.

The short American Conservative article doesn’t mention the taxation of capital, but I wouldn’t be surprised if George mentions this in his book on deflation, as it seems to me to be the best argument for the superiority of George’s proposal to mine.  Our current system taxes nominal returns on capital, not real returns.  More importantly, it may not be feasible to fully index the tax system for inflation, as with money constantly flowing in and out of various saving accounts it would be an administrative nightmare.  This means that we are probably stuck with nominal taxation of income from investments.  If so, then the lower the inflation rate, the smaller the distortion (remember that even taxing real capital income is a distortion, as an optimal tax regime would only tax consumption.)  In my view, this is the biggest potential benefit from mild deflation.  (If we could get a Singapore-style tax system, which exempts income from capital, then we would not face this problem.)

The issue that I see as being most important right now is macroeconomic stability, and in this area I believe that either my proposal or George’s would dramatically outperform inflation targeting.  Under inflation targeting nominal incomes rise whenever the economy booms.  But nominal wages are sticky in the short run, so these booms will be associated with high profit rates and macroeconomic instability.  Almost inevitably, they will be followed by at least a mild recession.  I think George’s intuition about what went wrong in recent decades with monetary policy is roughly correct.  Inflation targeting hid the buildup of monetary disequilibrium, or instability.  However his description of what went wrong focuses (in my view) too much on the distinction between inflation and deflation, and not enough on the distinction between inflation and nominal income growth.  Even under my proposal, the inflation rate would average about two percent a year.  And yet under my proposal we wouldn’t have the boom/bust cycle of rapid NGDP growth followed by a sharp slowdown in nominal GDP growth.

I am not certain exactly how George sees this issue, but the following quotation from his web page suggests that at times he focuses on the distinction between inflation and nominal income:

But a stable price level is far from being ideal in an economy with changing productivity: it does not minimize the burden borne by the price system, and it does not contribute to the efficient working of fixed nominal contracts.  The reason, in a nutshell, is that it is the stability of nominal spending (domestic final demand), and not that of the price level per se, that is crucial to general macroeconomic stability.  When productivity stays constant, zero inflation is equivalent to constant final demand, but not otherwise.  Instead, stability of final demand requires that a rate of deflation equal to minus the rate of productivity growth.  I call a monetary target based on this rule a “productivity norm.”

In my view, the problems that George refers to in this quotation can be addressed by moving to a NGDP rule, even if it allows for positive inflation.  However, I don’t think George sees things that way, as in an earlier email to me he expressed the view that a policy of positive inflation could create macro instability, even if linked to a NGDP target.

Although my mild inflation preference at first seems far removed from George’s mild deflation preference, I actually think we are very close on the single most important issue, that nominal income targeting is the key to macroeconomic stability.  I still intend to read George’s book, and then will have a better idea of the issues that separate us.  In the meantime, this gives you some preliminary sense of the issues involved.  I think that it is actually a pretty close call as to which approach is best.  If one goes by the old “Golden Mean” maxim, one might end up proposing a 3% nominal income rule (as frequent contributor Bill Woolsey favors.)  As far as I know, not many prominent economists favor any form of NGDP targeting—I believe that Bennett McCallum is the most well known advocate of this policy.

Please let me know if I have missed any important aspects of this debate.



19 Responses to “George Selgin on Deflation and Nominal Income Rules”

  1. Gravatar of Alex Golubev Alex Golubev
    12. March 2009 at 12:28

    until we fix bank portfolios, they won’t lend as much as is needed. Sounds like we didn’t need the banks in the 30’s. The RFC seems to have adressed the constraint of velocity of money that was being impeded by sick banks. Please comment (excerpt from a blog):
    “”…reports of the Comptroller of the Currency tells us that the banks sat out the recovery. The summary Table 47, “Total Assets and Liabilities of National Banks, June 1933-June 1937,” Report of the Comptroller of the Currency 1937 (Washington: GPO, 1938), pp. 488-94, shows that during the recovery, [banks’] total deposits increased 52 percent, holdings of government securities increased 57 percent, and (idle) reserves held with Federal Reserve banks increased 140 percent. Meanwhile loans and discounts–that is, the extension of credit to businesses and/or consumers–increased only 8 percent.

    So there was an economic recovery from the depths of the Great Depression with no financial fix, or rather with almost no participation by the banks, except of course that they bought the government securities that financed net contributions to consumer expenditures out of federal deficits. And no nationalization, either. How is that possible?

    In short: the Reconstruction Finance Corporation replaced the banking system as the lender of first resort to businesses large (the railroads) and small (most of its loans were under $100,000). The volume of its loans and discounts during the four years of recovery was roughly five times that of the national banks.

    The way to deal with the current crisis short of nationalization may, then, be to bypass the moribund banking system with a new RFC capitalized with the remainder of the TARP and other funds authorized by Congress.”

  2. Gravatar of ssumner ssumner
    12. March 2009 at 12:57

    Alex, The RFC dispersed 1.5 billion in 1932, 1.8 billion in 1933, and 1.8 billion in 1934, before sharply curtailing its lending. I see no evidence that it played a major role in the dramatic turnaround after March 1933. On the other hand, the value of the dollar was sharply correlated with the recovery, even if you look at daily data correlating the dollar with stocks, risk spreads, commodities, etc.

    I agree that the banking system is very important for a healthy economy. But even if the banking system is not working well, AD shocks can have a big short term imact on output. The main problem faced by manufacturers today is not financing, it is demand. If the orders were pouring in most of them would find some way to fill those orders. But the orders are not pouring in.

    And if we have more AD, the balance sheets of banks will improve dramatically.

  3. Gravatar of Bill Woolsey Bill Woolsey
    12. March 2009 at 13:24

    I guess I am just a traditionalist, but I think 3% growth in expenditure is just about right. I think nominal incomes should rise to reflect growing real incomes and the price level should remain pretty much stable. If productivity grows extra fast, yes, there would be mild deflation. And slower than usual real growth would be inflationary.

    However, I don’t favor moving from the Fed’s 2% inflation target (5% nominal income growth target) in the midst of a banking crisis. A disinflationary regime shift is going to cause problems at the best of times. No need to cause any more bank failures.

    Similarly, I oppose government deposit insurance. But don’t propose going cold turkey in the middle of a banking crisis. Right now, FDIC resolution is the best way we have of dealing with insolvent banks.

    If using open market operations to target the nominal Federal Funds rate is inconsistent with price stability, then a new operating procedure needs to be adopted. If it was true that we could continue with the great moderation forever due to the wonders to the Taylor rule with a background of 2% inflation, then maybe it would be worth it. But, that was wrong. At the very least, a back up is needed. When the target hits a negative rate equal to the cost of storing currency, then you have to go with quantitative easing. So, why not price stability as a target?

    I think the reason why nominal wages are sticky in a downward direction is that usually they signal a need to shift labor out the a sector with reduced demand. If is really true that those in the sector have no other opportunities and so all should just suffer a reduced standard of living, then it should be a big deal and not something that is sneakily implemented by a fall in the purchasing power of money.

    Generally, with more modest shifts, slower growth in real and nominal wages in shrinking sectors and more rapid nominal and real wage growth in growing sectors–seems just about right.

    On the other hand, when there is no need to redeploy labor, say, because there is an increase in the demand for money or a decrease in the supply of money, this usually sensible instituion has undesirable effects. Generally, I think aggregate supply shocks, changes in terms of trade, work out better with changes in the price level with nominal incomes growing along on trend. Generally these changes do require redeployment of labor and other resources, but there is also a change in real wages in general that don’t signal a redeployment. A temporary or permanante reduction (or increase sometimes) in the standard of living must be asorbed by everyone.

    Have you ever considered that the reason why the Fed is not inclined to make specific targets is because of politics? Maybe money illusion is most effective at the political level, and committing to inflaition might be politically unpopular? That explaining that we need inflation so that your real wages will be depressed without much fuss might not be appealing?

    I grant that 5% total spending growth might not raise the red flags that 2% inflation might raise. But if the goal is a strong committment to a nominal target, then it is something to consider.

    P.S. Unless Selgin has changed his views a lot, he supports free banking with no central bank, so “central bank” operating procedures are irrelevant. He argues that any growth in nominal income requires a market interest rate below the natural interest rate that creates malinvestments. In the long run, the interest rates must readjust and the malinvestments are liquidated.

  4. Gravatar of Alex Golubev Alex Golubev
    12. March 2009 at 13:26

    But AD is a function of the availability of credit, no? or am i speaking from a different school of economics, if i say that? I think availability of easy credit and home ATM’s had something to do with the granite countertops and hummers that people were buying. So if we extend that logic, removing that credit, would first directly reduce AD and then influence people’s confidence which would further reduce AD. so to the extent that RFC provided credit where the banks failed, it stimulated AD. What i’m trying to say is that RFC increased the velocity of money which in turn raised the stock of money/credit and reduced the value of the dollar. I’d love to check out the value of the dollar back in the 30’s but i don’t have that data. i have no doubt that it IS correlated to everything, since it IS the denominator after all. that’s why i don’t think we really disagree.

  5. Gravatar of Winton Bates Winton Bates
    12. March 2009 at 15:00

    The question of whether there is a bias in measurement of inflation is also relevant. I recall a few years ago some researchers suggesting an upward bias of around one percent, but I’m not up to date with research findings in this area.

  6. Gravatar of travis travis
    12. March 2009 at 15:03

    Not really relevant to the post, but I thought that you should know that Britain is doing quantitative easing in a big way right now:

    They announced this policy on March 5. I don’t think that it is a coincidence that stock markets have rallied dramatically since then.

    Lately, the Fed often follows what Britain does. It will give Bernanke leverage to beat up on the Fed governors who are holding out on him and preventing a near unanimous move to greater easing.

    Let’s hope that Bernanke cuts the interest rate on reserves.

  7. Gravatar of Bill Woolsey Bill Woolsey
    12. March 2009 at 15:03


    I think you are coming form another school of economics.

    Credit involves both a borrower and a lender. While it allows the borrower to spend, the lender has less money to spend.

    In the back of your model is that money holdings passively adjust to match, ‘not lending,” or “not spending.”

    So, if there is less credit available for home imrpovements, those who don’t make the loans are assumed to have simply hoarded “money.”

    Now, it is possible that people can choose to hold more or less money. And, it is possible that people will choose to hold less money and more RFC debt. And so this could raise spending.

    But it is the medium of exchange that gets spent. People who have less money than the want to hold can simply spend less out of their income and accumulate it. And peopel who want to get rid of money don’t have to sell it to someone who wnats to hold it. They just spend it.

    It is because money is generally accepted in exchange. It isn’t an investment asset.

    Again, that isn’t to say that people don’t hold wealth in the form of money and it can change.

    And, anyway, there is a budget constraint that it begin ignored if you just add up different kinds of spending to get aggregate demand. Or assume that credit generates spending without thinking about what the creditors would have done with the money instead.

    Always ask the question.. where did the money come from, and what would they have done with it?

  8. Gravatar of David Beckworth David Beckworth
    12. March 2009 at 19:14


    Another interesting post. A couple of thoughts. First, downward nominal wage rigidity might be the result of living in a world where inflation is the norm. What if inflation were not the norm? Christopher Haynes and John James in the AER (2003) find no evidence of downward wage rigidity during the secular deflation of the postbellum period. But even if nominal wages were rigid in a “good” deflationary environment would they really be bothersome since real wages would be increasing?

    Second, under Selgin’s scenario I am not sure that nominal interest rate would necessarily get close to zero. The good deflation is coming from productivity gains which implies, all else equal, a rise in real interest rates. This rise in the real interest rates would provide an offset in the nominal interest rate to the downward pressure created from the deflation.

    The key, though, is that both you and Selgin see the importance of stabilizing nominal spending. Had this been either in your or Selgin’s form, I suspect many of the current problems would have been avoided. I make this point in a recent article where I look at the good form of deflation (

  9. Gravatar of Jon Jon
    12. March 2009 at 19:57

    “They announced this policy on March 5. I don’t think that it is a coincidence that stock markets have rallied dramatically since then.”

    As I’ve mentioned on other threads, this is actually what the US Fed does normally but oddly has not done in the past year. Open market operations are constructed as purchases of notes and bonds. The counter-parties are “primary dealers” such as GS not depository institutions. PD act as intermediaries vis-a-vis the public. For more, look at my earlier comments.

  10. Gravatar of ssumner ssumner
    13. March 2009 at 05:26

    Bill, I could probably be talked into 3% after this crisis is over if I could be convinced that the Fed was moving beyond their exclusive focus on short term rates, which has been disastrous in this environment. Remember, the target rate fell to 1% in the last recession ,what would it have been if we entered that recession with a trend rate 2% lower than it actually was? So I think your plan would be feasible in an environment with good monetary policy, indeed might be optimal, but I am still a bit risk averse about taking the plunge until I have more confidence in the Fed. Think how astounding it is that we stumbled into a liquidity trap with a trend rate of inflation 2% above the gold standard trend, and they rarely had liquidity traps under the GS!

    I see the wages and money illusion concept a bit differently. If money illusion exists around the boundary between rising and falling nominal wages, then a slightly higher trend rate of inflation would greatly reduce the number of times workers decisions were (irrationally) distorted by that boundary. I want them to make more rational decisions, not be fooled. If there is a trend rate of inflation, then they will make decisions about when to switch into other industries based on whose getting the biggest pay increase, but they won’t perversely spite themselves into unemployment because they fell a 1% nominal wage cut is losing face. (When their real wage might have been rising.) And when you talk about times when labor doesn’t need to be redeployed, remember that there is always some churning in a large complex economy, so some workers will always see relative pay cuts. Politics may explain the lack of an explicit inflation target, but some central banks have done it, so the political resistance probably is not insurmountable.
    One good argument against my plan is that if 5% NGDP becomes enshrined in policy, then the public may begin to expect 4% pay raises as the norm (assuming one percent population growth) and then anything less might be viewed as just as outrageous as a nominal cut is now viewed.

    Alex (and Bill’s response), I am willing to let you use whatever model you want. If you want to use the Keynesian expenditure approach, that’s fine. I’m a pragmatist who believes in looking at the evidence, not fitting facts into my preferred, pre-concieved theories. [As an aside, I would much rather that money illusion didn’t exists, as you guys know I am a big Ratex, EMH supporter, but money illusion does seem to exist (maybe.)] So I would say let’s look at what happened in 1932, when the RFC was already in full swing. In 1932 we saw one of the largest declines in base velocity, as prices and output fell to depression lows. So I see no evidence that it boosted velocity. Velocity didn’t even rise in early 1933, it rose precisely when the dollar started being devalued, and the rise in velocity (as far as we can tell) closely tracked that devaluation.

    Winton, I agree that bias exists in price indices, indeed there really is no perfect index, as assumptions always must be made about weighting. But these biases don’t impact NGDP, which is the preferred target of George and I. The bias affects the estimated inflation rate that would result from each NGDP rule.

    Travis, That observation is very interesting and we need to follow this issue closely. My unorthodox approach to monetary indicators (which says interest rates and the money supply may be a misleading indicator of policy) makes it hard to identify policy shifts. Indeed I have always thought many important shifts occur unintentionally, as the (unobservable) Wicksellian natural rate creeps above the policy rate. If the Fed just hints that it might move toward more QE, that could slightly boost inflationary expectations, slightly boost the Wicksellian natural rate, and make the current zero rate policy somewhat more expansionary in practice. I know all of this sounds frustratingly vague, but I just don’t see any shortcuts, any perfect indicators of what’s going on. That’s why I look at the markets (where I think commodity prices bumped up a bit as well.) So I’d say I am still somewhat pessimistic overall (and think more should be done) but am also a bit more optimistic than a week age. Rapid recoveries from AD shocks are possible (more so than AS shocks), and I wouldn’t rule out an upside surprise. Even though the Fed has lost some credibility, I still think they have more anti-deflation credibility than the BOJ. We should never have had the current collapse, but I still think we will avoid having the GDP deflator fall 10-15% over the next decade, as happened in Japan. Expectations and wages will adjust, and may allow a decent recovery, unless Fed policy gets even more deflationary than I now think likely. So there is hope.

    David, Your first point about the experience under the gold standard is a very good one, and if true would certainly weaken my argument. I’d want to see evidence that the distribution of wage changes did not show a strong discontinuity around the zero percent point. But I don’t agree with your second and third point:

    1. Although real wages would usually be rising under a Selgin-type system, they would rise at different rates. That means that if the average nominal increase was zero, then some of those real wages rises would be associated with falling nominal wages. So if I am right about money illusion, then the fact that the overall real wage rate rises due to productivity doesn’t eliminate the problem. So I think you should put more weight on your first argument–about the 19th century evidence.

    2. On the real interest rate question, I think you either misunderstand the issue, or more likely understand it but overestimate its empirical relevance. Productivity growth does lead to positive real interest rates, and the higher the productivity growth the higher the real interest rate. So far, so good. But there are an incredible number of factors that affect the long run productivity growth rate of an economy. (Technology, science, human capital, education, an efficient tax system, efficient regulations, rule of law, etc. etc. The difference in the long run productivity under my system vs. George’s preferred system is likely to be small, EVEN IF GEORGE IS RIGHT. The current rate is around 2%, and in my view any difference between the two systems would be a very small fraction of that number. So even if George is right about the superiority of his system, I don’t think any fair-minded person could estimate it’s impact on productivity growth (and hence real interest rates) as exceeding a couple tenths of a percent. So the 4% inflation differential still looms very large (through the Fisher equation.)

    Jon, Yes, I think I agree with you. I think you are saying that all the various facilities recently added to the Fed’s balance sheet are not for QE, but to either bailout or provide liquidity to the banking system. And that if they are serious about QE, we should see more ordinary OMOs. And this also gets back to my observation about recent trends in the base. Bill checked and confirmed my observation that the base has recently fallen sharply. I take that as the Fed feeling the banking system is getting a bit more stable, and doesn’t “need” as much reserves. But as long as that “accommodation” mindset is at work, we still haven’t begun serious QE, which is trying to put more money out there than the banking system “needs.”

  11. Gravatar of David Beckworth David Beckworth
    13. March 2009 at 07:13


    Thanks for the reply. Regarding the real interest rate my point was not that productivity would be higher under Selgin’s view. Rather, it was that the increase in the real interest rate coming from productivity–whatever its sources–would tend to offset the downward pull of deflation on the nominal interest rate. In short, both terms in the fisher equation would change–the real rate would go up, the expected inflation part would go down–making it less likely of hitting a zero nominal interest rate.

  12. Gravatar of Lorenzo (from downunder) Lorenzo (from downunder)
    13. March 2009 at 16:28

    Speaking as (small) business owner who employs people, the reason why wages are sticky downwards is that one is in a continuing relationship with one’s employees. It is not worth screwing around with that relationship to turn wages into some sort of “spot price”. Even when the business, as the one I am part-owner of does (we put on medieval and ancient days for schools), operates on an “offer and accept” basis with a roster of casual employees and highly variable demand from week to week. It is one thing to vary how many full or half-days work we offer each presenter, it is another to vary what we actually pay them for what we offer, given labour is normally purchased as on-going connections.

  13. Gravatar of Lorenzo (from downunder) Lorenzo (from downunder)
    13. March 2009 at 16:37

    And, I should add, they have a particular pattern of expenditures to maintain. I guess, in some ways, wages being sticky downward is the other side of the permanent income hypothesis.

  14. Gravatar of ssumner ssumner
    13. March 2009 at 17:43

    David, I still think that if you lowered the average inflation rate from about 2% to about negative 2%, the average nominal rate would have to be much lower. But I do agree with you that if productivity is robust, you can have positive interest rates even with deflation. I just think there would be more situations where rates are near zero, as risk free real rates are often around 2%.

    Lorenzo, I agree with what you say, I would just add one point. I think a lot of people in the real world look at things a bit differently than economists. We think in terms or real wages as being all that matters. So we think it should be just as insulting to give workers a 5% raise when inflation is 7% as it is to give workers a 2% pay cut when inflation is zero. But in the real world the latter seems worse (I think.) I’d be interested in your perception. Do you agree that the workers would view the second case as “screwing around with them” more than the first example.

  15. Gravatar of Bill Woolsey Bill Woolsey
    14. March 2009 at 05:40

    There have been substantial wage cuts in the current recession. There was an article about this being more or less the encouraged policy in S. Korea. The articles in the U.S. have focused on management.

    I actually heard a union flack explaining that avoided wage cuts and raising wages are what we need to raise spending, so the economy will recover.

    Anyway, my view is that if demand for the product is low, the wage custs are completely feasible. Discuss it with all the workers together. If some workers want to say, let the slackers go… fine. If tehy say, we just want to take more time off, so reduce hours but pay us at the same rate.. fine. When the opportunity to dampen that effect by cutting back the work time less but making at lower wage rate, leaving total wage income higher.. I think that will be a reasonable option.

    So, lowering nominal wages as a way of firms cutting losses and dampening output drops is entirely reasonable.

    For expanding firms, they can always hire new workers for less than existing workers.

    Anyway, protecting incumbent workers in profitable expanding firms from the nominal wage cuts doesn’t seem to me to create insurmountable barriers to using wage cuts to clear labor markets in surplus.

  16. Gravatar of ssumner ssumner
    15. March 2009 at 05:49

    Bill a couple observations on wages cuts.

    In an early post I showed that FDR’s high wage rate low hours policy was disastrous. This supports your point about low wages high hours.

    Wage cuts could help boost employment in a recession, but there are two problems with relying on it as a macro policy.

    a) It probably isn’t practical, and thus will appear to fail even if tried. So the government shouldn’t waste time encouraging wage cuts, but rather should get the same effect by boosting NGDP growth up to the target.

    b) Recessions have strong sectoral impacts, hitting certain types of manufacturing (capital goods, etc) very hard. That’s why East Asia is suffering, despite their relatively good banking record during the subprime crisis. If the entire economy needed a 5% wage cut to restore full employment, and if most healthy industries did not cut wages at all, then even a 20% wage cut in highly cyclical industries might be woefully inadequate to restore full employment. So again, I think wage cuts may be needed in certain situations, but they are not a good government policy.

    In defense of wage cuts I would make the following observations. The 1920-21 depression might be the best example of wage cuts working. The U.S. was still firmly committed to the gold standard, which required some deflation after the high prices of WWI. Because wages fell rapidly, the severe recession was followed by a very quick recovery. In contrast, in 1929-30 Hoover strongly pressured industry not to cut wages, and (for whatever reason) wage cuts were indeed much more slow in coming than in the 1920-21 case. And we all know how Hoover’s policy turned out.

    So I don’t think the government should either encourage or discourage wage cuts.

  17. Gravatar of W.E. Heasley W.E. Heasley
    28. March 2009 at 12:32

    The Fed has basically targeted inflation by the use of increasing or decreasing short term interest rates. However, in an environment of a nearly zero interest rate, the Fed’s mechanism of interest rate policy for dealing with inflation has reached its limit.

    Nominal Income targeting might well be an alternative to interest rate targeting policy regarding inflation.

    Regarding Real Income and Nominal Income, the average consumer does seem to confuse the two subjects except in the cases of Taxation and during periods of high inflation with rapidly changing prices.

    Consumers seem to have a very clear understanding of marginal income tax rates and the relation of taxation to Real Income.

    Also, during the 1970’s with hyper inflation, consumer were very aware of Real Income. Matter-of-fact, during the 1970’s the Press ran stories on a constant basis regarding Real Income as Inflation was a headline item. The only place I can think of in the Press that reports Real Income growing in the recent 4 to 6 months is The Kudlow Report. Larry Kudlow reports on Real Income at least once a week.

  18. Gravatar of ssumner ssumner
    29. March 2009 at 10:02

    W.E., I mostly agree. One point you need to be aware of, however. The fed funds rate is a policy instrument, a means to an end. NGDP is the policy goal, the end it itself. You can use short term rates to influence NGDP, or you can use the monetary base, or exchange rates. Those are all policy instruments.

  19. Gravatar of Deflation, Rothbard’s critique of Friedman « Anthony Hennen Deflation, Rothbard’s critique of Friedman « Anthony Hennen
    12. May 2011 at 16:44

    […] see this Cato Journal article, a blog by David Beckworth, and two blogs by Scott Sumner ( here and […]

Leave a Reply