Thursday, February 5, 2015

Bachmann, Berg and Sims on inflation as stimulus

RĂ¼diger Bachmann, Tim Berg, and Eric Sims have an interesting article, "Inflation Expectations and Readiness to Spend: Cross-Sectional Evidence" in the American Economic Journal: Economic Policy.

Many macroeconomists have advocated deliberate, expected inflation to "stimulate" the economy while interest rates are stuck at the lower bound. The idea is that higher expected inflation amounts to a lower real interest rate. This lower rate encourages people to spend today rather than to save, which, the story goes, will raise today's level of output and employment.

As usual in macroeconomics, measuring this effect is hard. There are few zero-bound observations, fewer still with substantial variation in expected inflation.  And as always in macro it's hard to tell causation from correlation, supply from demand, because from despite of any small inflation-output correlation we see.

This paper is an interesting part of the movement that uses microeconomic observations to illuminate such macroeconomic questions, and also a very interesting use of survey data. Bachman, Berg, and Sims look at survey data from the University of Michigan. This survey asks about spending plans and inflation expectations. Thus, looking across people at a given moment in time, Bachman, Berg, and Sims ask whether people who think there is going to be a lot more inflation are also people who are planning to spend a lot more. (Whether more "spending" causes more GDP is separate question.)

The answer is... No. Not at all. There is just no correlation between people's expectations of inflation and their plans to spend money.

In a sense that's not too surprising. The intertemporal substitution relation -- expected consumption growth = elasticity times expected real interest rate -- has been very unreliable in macro and micro data for decades. That hasn't stopped it from being the center of much macroeconomics and the article of faith in policy prescriptions for stimulus. But fresh reminders of its instability are welcome.

At first blush, this just seems great. Finally, micro data are illuminating macro questions.

It's cleaner than the  Hagedorn, Manovskii and Mitman paper I blogged last week, because many of the aggregation issues are absent. There, I complained that employment in one state might be  gained by business moving from another, which would not be an available channel for the whole economy. Here, if we know that people who expect more inflation spend more, it's an easier jump that if we all expect more inflation we all want to spend more. This aggregation problem is usually one of the biggest stumbling blocks for the project to measure macro effects from micro data.

Now, for a little whining. This isn't really criticism as I don't know how to do any better. But it does make for a very well-done example in which to ponder the limitations of the micro evidence on macro questions methodology.

Here are Table 1 and 2, the "baseline specification."

It's a probit regression. The left hand variable is whether a person answered yes or no to the question,
Q1: “About the big things people buy for their homes—such as furniture, a refrigerator, stove, television, and things like that. Generally speaking, do you think now is a good or a bad time for people to buy major household items?” 
The main right hand variable, ("Inflation expectations (1Y)") is the answer to the question,
Q2: “By about what percent do you expect future prices to go (up/down) on the average, during the next 12 months?”
The main fact is that the top row of numbers are all essentially zero, decently well measured, and nonetheless statistically insignificant. Where it is significant, in the zero-bound years, it's negative -- higher inflation expectations are associated with plans to spend less, not more!

So far, so good. But what are all those other numbers in the table? Well, these are "controls," extra right hand variables in the regression.

What in the world are they doing there? The fact is not "people with higher inflation expectations don't plan to spend any less." The fact is that "people with higher inflation expectations, holding constant their expected financial situation and income, their expected change in nominal interest rate and aggregate business conditions, ..., a long vector of aggregate variables, and then the whole Table 2 of demographic variables, don't plan to spend any less." Hmm.

The long list of "controls" brings back memories of all the regression horror stories I was taught in graduate school (thank you Tom Rothenberg).

Left shoe sales = a + b price + c right shoe sales + error. 

Wage = a + b education + c industry + error. 

(In case the latter isn't obvious: including industry helps a lot to "explain" wages and raise R2. But the point of education is to let you change industries from fast food to computers, so you absolutely do not want to "control" for industry!)

What are all the controls doing here? Could we not at least start with OLS, a clean digestible fact, or a graph so that poor bloggers have something to brighten up posts?

I asked the correspondent who sent me the paper (thanks) who opined that the referees probably made the authors do it, and out of a reasonable concern. Maybe the correlation between inflation expectations and spending plans across people does not measure the causal effect, what if we change inflation and leave other things constant?  It could well be that the correlation of expectations across people is zero, reflecting other forces at work, but if we raise everyone's inflation expectations, then we would raise everyone's spending.

Most simply, just because we put inflation expectations on the right hand side of a regression and spending on the left, does not mean that changes in inflation expectations across people cause their spending plans to change.

Demographic controls seem reasonable. Suppose the fact was that women all expected higher inflation and planned to spend a lot, while men expected low inflation and did not plan to spend a lot. One would not want to use that correlation to measure how increasing expected inflation for all of us would affect our spending. Such a demographic correlation is much more likely a result of other causes affecting both variables (inflation expectations and spending). This really remains the deep issue of micro to macro implications: Does a correlation across people tell us what happens if something affects all of us?

But if demographic controls changed the result a lot over OLS, one would be very suspicious. A correlation that survives controls is a lot more persuasive than a correlation that only emerges with controls. It's much nicer to say there is a raw correlation, and verify that it is not the result of differences between demographic groups, than to say the correlation is only measured after demographic controls. Because no set of controls is perfect. (The implicit assumption "my controls perfectly capture all the reverse causation or all third variable influences" pervades regression analysis.)

Many of the controls are macro variables. There are almost as many controls here as time data points. Year dummies would have removed all the time-series variation and left us the pure cross section a lot more simply.

The first set of controls for other expectations strikes me as the most fishy. Why would we measure the effect of a change in expected inflation holding constant expected unemployment? The whole point of the macro experiment is to raise both expected inflation and to lower expected unemployment.

This is the hard nut of all regression analysis: why does the right hand variable vary? People spend a lot of effort on the left hand variable, but that's actually less important. What caused the variation in your data? We don't have randomized experiments. Why is it that households have such widely (insanely!) varying expectations of inflation? Until we know that, it's really going to be hard to tell whether their similarly widely varying spending plans are because of higher inflation expectations, or because inflation and spending plans are both results of some third cause.

The paper isn't much help on this issue. At least I wish they (or much of any regression work) at least asked the question. They don't even really discuss the "controls" in this way; why expected inflation varies, and then control for determinants of expected inflation that are correlated with determinants of spending.

The discussion of the control variables sounds a lot like the habit of assuming everything on the right is a "cause," and fishing for R2, like left shoes in the right shoe equation, and industry in the wage equations.
With respect to the coefficients on the economic control variables, we obtain for the most part plausible and significant estimates,... the expected financial situation of the household and its real income, the expected business conditions (idiosyncratic and aggregate), the current financial situation, and the current real household income all have significantly positive effects on the reported spending readiness. In addition, a positive judgement of US economic policy also affects spending dispositions positively. Moreover, an expected increase in future nominal interest rates makes people want to spend more today,  while higher economic uncertainty in the form of stock market volatility, inflation volatility and higher unemployment rates (both current and expected) decrease the probability that people find buying conditions favorable ...
But enough whining. My point is that micro, regression-based analysis has its limitations too. This seemed like a good example on which to remind graduate student readers of common regression pitfalls: Always ask what caused the variation in the right hand variable. Use minimal controls, not the kitchen sink. Make sure the partial effects of your regression (move x holding z constant) make sense. And so on.

But I don't think I could have done better, as making sense of why people's expectations are as widely dispersed as they are seems a big challenge.

It's still a powerful observation, and I trust it's there in the OLS with minimal controls. People who expect more inflation do not plan to spend more. If you think raising all our expected inflation will make us all spend more, you have some creative explaining to do.

Update: Eric responds:
On your point about all the control variables . . . we did (more or less) what you suggest in the blog post. If you look at Table 3, we drop all of the idiosyncratic control variables in one specification and get essentially the same results; also in Table 3 we do the version with time fixed effects instead of aggregate controls. If you go to the online appendix, in Table 8 we show raw correlations between expected inflation and buying attitudes. We also split the raw correlation by a large number of different demographics. In Figure 7 we show plots of time-varying raw correlations between expected inflation and spending attitudes -- it is the analog of Figure 6 in the main paper which plots a time-varying marginal effect based on the probit estimation. Basically this all shows exactly what you ask for in the blog post -- the correlation/coefficient between expected inflation and buying attitudes does not depend on the controls.
I admit not reading all the way through or the online appendix. They also confirm that the early drafts started with raw correlations. There is an interesting writing (and editing and refereeing) conundrum, should a paper start with the "main" result, or should one start with suggestive robust facts and correlations and then address objections with a more sophisticated model. It's not an easy question -- Most papers drag you through 10 tables of motivation and summary statistics and suggestive correlations before getting to the point, and I really admire that this paper had the main result on Table 1.  OTOH, by going the other way around busy bloggers miss the interesting correlations in online appendix Table 8!


  1. What was the actual inflation question asked? I wonder how many people are conspiracy nuts who insist inflation is north of 20% (and has been for years)? I know some and they seem to live in this imaginary world while making spending decisions in the real world that reflect actual sane beliefs rather than their stated beliefs. They act on the stated belief by buying gold, guns, and ammunition, but not for durable goods.

    1. Ben, the actual question does not mention the word "inflation" but instead asks about percent change in prices.

    2. That doesn't change the problem. These guys know that inflation means % change in prices. They cherry pick a few prices with high percentage changes and allege other prices must be as bad.

  2. It seems the issue is not whether more expected inflation stimulus consumer spending. Raising expected inflation in the late-1970s did not benefit the economy.

    Rather, the issue is whether a lower real interest rate stimulates spending. Not the same thing. With nominal rates at the zero bound, higher expected inflation might make a difference.

    Moreover, consumer spending isn't so much of an issue. Investment spending is where the rubber meets the road as far as the impact of interest rates matters. Would lower real interest rates boost investment?

    Finally, the whole issue seems a distraction. New Keynesians emphasize the liquidity trap. Older Keynesians emphasize price and wage rigidity. The prices of financial assets traded in centralized markets move rapidly. But prices of heterogeneous goods in decentralized markets do not. Nominal wages and nominal house prices move slowly to their new equilibrium. If they are too high, higher CPI inflation is a way to grease the wheels of adjustment, allowing a fall in real wages and real house prices via a rising CPI.

    I think if we had higher CPI inflation in the past six years, the recovery would be more advanced.

    1. Is it not the case that real interest rates reflect the return to invested capital, i.e the marginal product of capital? If real rates (mpk) are negative, why would we see higher investment spending? Real rates were high in the 1980s and investment spending boomed.

  3. That may have been an idea years ago, but today higher inflation is advocated because it lowers the average debt burden in real times, preventing further demand shocks from further deleveraging/defaults if peoples private debt were to continue to accelerate. Public government debt also needs to be lowered or stabalized in real terms by bringing NGDP back to trend, preventing further demand shocks from high interest rates or heavy austerity. I don't think it's advocated inherently due to any expectations channel, inflation is more spending, people are just adovacting higher nominal spending and higher nominal income. They're advocating the result.

  4. Good piece on an interesting paper! Questionnaires, however, always make me a bit uncomfortable. Here, the questions are, probably in order to avoid framing accusations, deliberately general. E.g., Q1 is not pertaining to the individual's own spending plans, but to "people's". And Q2 is unclear on which prices are meant to increasing or decreasing. A clean test of intertemporal substitution would be to have asked along the lines: Q1: Do you plan to buy durables this year? Q2 If "Yes to Q1", would you buy more or less durables if durables gets more expensive next year?". But maybe these would be considered leading questions? But in any case, when talking about intertemporal substitution in DSGE models we are interested in how much "more" is, not whether individuals think others should spend or not. Nor whether they want to spend themselves or not. It is _when_ a given spending occurs that is at core.

  5. I'm a little surprised, John. You are usually crisp and sensible on issues like, "does a higher minimum wage reduce employment" or "do extended unemployment benefits reduce employment." The answer you (and I) would give is, "of course."

    Then why is it not obvious that increasing the cost of holding money will cause people to hold less of it at the margin?

    Higher inflation expectation will make people want to hold less cash, which will make them look harder for somewhere to invest it, which will cause someone else to spend more. Heating up the potatoes will increase the hot potato effect, surveys or no surveys.

    The fact that some survey finds no correlation between people's stated inflation expectation and stated propensity to spend sheds no light whatsoever on the question of whether a higher inflation target would increase output when we are facing an output gap and a ZLB.

    Kenneth Duda
    Menlo Park, CA

    1. Yes, if you pay for something you get more of it. But how much is an open question. Just how much extra or less saving does a one to two percentage point greater or lesser real return, for a year or two, actually induce? If you think about the utility costs of not adjusting savings habits quickly to interest rate changes they aren't all that big. And as a finance person, I think the risk premium is much more important for saving/investment rather than the real interest rate on short term bills. Inflation is also poorly measured down at the last 1-2 percentage point. 10 or 20 percent inflation, and I think you'd see some pretty big responses!

    2. Just so my plan based upon the Chicago Plan of 1933 and Fisher's debt-deflation ideas is clear, inflation expectations do not influence consumer spending, for example, but investment in the stock market and bonds. Here's what Pimco says about that:

      "Understanding inflation is crucial to investing because inflation can reduce the value of investment returns. Inflation affects all aspects of the economy, from consumer spending, business investment and employment rates to government programs, tax policies, and interest rates."

      The monetary policy/QE is what is intended to raise inflation expectations and incentivize investment sooner rather than later, especially the riskier investment that creates new businesses and jobs and can protect investors from higher inflation. Just so, keeping low yields on shorter term investments is a disincentive to invest shorter term if longer term rates are rising.

      The govt borrowing/stimulus is what is used to influence current spending in the economy. A dated coupon, payroll tax holiday, sales tax holiday for states, cash for those who are suffering through the economic downturn, etc., is what are used for increased current spending.

      The govt borrowing/stimulus is also used to reinforce investment through tax credits for investment, a payroll tax holiday to lower wages, a credit for hiring, etc. The so-called infrastructure spending also influences current spending and investment.

      The govt borrowing/ spending also reinforces forward inflation expectations which the monetary/QE is mainly addressing. The govt borrowing follows from lower tax revenues during an economic downturn, and the desire to neither raise taxes nor severely reduce employment during such a downturn.

      As the economy improves, tax receipts will rise, the policies mentioned above will be phased out, lowering spending, and inflation will also help in lowering the govt debt. Over time, the inflation expectations will be due to the recovering economy heating up and not the govt borrowing.

      This all might well be wishful thinking, but I do feel some confidence that investors do pay attention to inflation expectations, even if consumers do not.

    3. Interesting. Thanks for the reply. All of your points make sense.


    4. I'm with Kenneth.

      The point of monetary stimulation is not necessarily to get final consumption up very much. It is more about getting people, businesses and banks to move their savings from cash to business investment and getting business spending up in order to enable future consumption, especially in a world with an oncoming wave of retirees.

  6. Can the argument also be transmuted into, if the market has a reasonable belief that prices will fall, then they'll hold money and spend less then they otherwise would have if they had expected prices to rise?

  7. The negative coefficients are indeed suspect. No question that high correlation among multiple coefficients can produce that.

    That said, ignoring all the controls, it's pretty clear that inflation should increase the trading of nominal assets for real assets, to the extent that prices of real assets have not already adjusted for the expected inflation. But, it's not clear what happens first: do prices adjust first, or does spending affect prices?

    We have this problem in my area of work, investments: Do stock prices adjust due to trading, or does the price adjust to information before people have a chance to trade? The microstructure literature is rite with models, and the conclusion is: sometimes yes, sometimes no.

    So, in the end, given a wide range of people's expectations and ways that businesses set prices, you'd expect an increase in spending because sometimes the prices are slow to adjust. That would explain the negative coefficient on the current inflation rate.

    Of course, the real question for policy makers should not be whether inflation boosts trading of nominal for real assets. It should be whether this trick can be used to boost long-run utility of wealth. Even if we can boost spending in the short run, this pretty clearly comes at the expense of spending later on. It works much like debt, which takes its toll on productivity later on.

    So, the success of this as a wealth-boosting mechanism depends on the FED's ability to spot market bottoms and tops, so they are essentially forcing people to "buy low and sell high" on real assets. The FED famously cannot spot bubbles, so they also cannot spot market bottoms, and, so all these mechanics cannot possibly work to boost people's wealth anyway.

  8. Isn't there (also) a problem here with that the durable goods mentioned might improve in terms of flow of benefits to offset the cost of waiting? It's perhaps a silly question, but I don't have the feeling that Q1 and Q2 are about the same thing. Q2 is about the average of prices over all goods and Q1 is about a subset of goods.

    Furthermore, given that these are durable goods what is the alternative that individuals are faced with? There are either those who already have a durable good, and those who don't. The first would have to give up a flow of services and would be imperfectly compensated for it if they sold it on the second-hand market (some premium will have to be paid by the sellers to the buyers, in case it isn't as good as it is all cracked up to be) when they decide to replace the good sooner rather than later. The second would also have to give up a flow of services by waiting. Who are the marginal consumers and how many are there likely to be?

    As you said at 10-20 percent there might be an effect, which brings me to the final point. What is the power of the test? Is it even mentioned in the paper (I currently don't have access)? If the effect is likely to be very small, is the test powerful enough to detect it given the data used? Somewhere I doubt it. What effect size can be ruled out at, say, 80 % power?

  9. It's not the fault of the structure of the paper that you didn't read the whole thing before writing a review.

    1. Well, I do plead guilty there. I don't read all the way through a paper, check the algebra, and read the online appendices before reviewing. But, unlike some of my fellow bloggers, that's why I don't insult people for being stupid morons when I disagree with the first 200 words, and I do email the authors to see if I've been fair. Unlike many other bloggers, I read some of it and post a bit of an opinion rather than just pass on interesting links. Given the press of time, that is likely to remain the quality standard on this blog. Caveat Emptor.

      Sad as my laziness is as a blogger, it is a good lesson for all wannabee writers. Your readers are busy. They do not slog through the whole paper. Your referees and editors are unlikely to read the whole paper either. Sorry, it's just a fact of life. So structure your papers with very busy readers in mind, knowing almost all will leave before the end. If you're slowly building to Table 16 of main results, they will all be gone when you reveal the punchline.

  10. Nice comments about the use of control variables. However, isn't the issue that cross-sectional correlations, or lack thereof, are not the same as time-series correlations. Assume the authors' findings are correct: people that expect more inflation than the average person expects don't plan to spend more than the average person plans to spend. Suppose, however, that a uniform increase in everyone's inflation expectations does cause a uniform increase in everyone's planned spending but their relative (to each other) inflation expectations and relative planned spending does not change. Then, the authors' findings of no *cross-sectional* correlation would still be consistent with correlation between (time-varying) mean inflation expectations and mean spending *across time*. Isn't "looking across people at a given moment in time" just looking at the wrong data to understand the behavior of the average person across time?

  11. I'm not sure that the survey is of much use. It demonstrates not what people actually do but rather what they think they will do. It's time period is 1984 until 2012, beginning just as significant inflation was ending and running through a period of low inflation. A useful survey would examine what people actually did during the years of the Great Inflation. Obviously much harder data to gather, but it would have the virtue of being grounded in reality. It would be interesting to see what people in say Germany are doing today in an era approaching Deflation.

  12. May I suggest the effect is non-linear? Inflation in the low single digits will probably not materially affect consumption or spending. Inflation in the double digits, probably will. Suffice to look at the experience of Latin America during periods of high inflation.

  13. Egads, of course if you tell people there will be inflation but that their incomes will be fixed, you may get a response of less planned spending.

    How about if you ask people if they will spend more if their nominal incomes increase even more quickly than inflation?

  14. Interesting thoughts on the recent trend to micro-panel evidence in Macro. My biggest issue is that it seems to ignore the feedback (general equilibrium) effects which are what makes Macro Macro, and not just Bigger Micro.

    Take for instance the Nakamura & Steinsson (2014) paper who look at regional variation in US military spending to 'estimate an "open economy relative multiplier" of approximately 1.5'. By ignoring general equilibrium they ignore that putting military spending into on region may mean enticing people to relocate from another region thereby reducing output in the other region. An effect they do not measure. General equilibrium would need to take this into account, but no micro-panel studies of macroeconomic issues, at least that I have seen, appear to consider the general equilibrium issues, that is the macroeconomic issues. They mostly just seem like tests of the microfoundations.

    Why do different people expect such different amount of expectation? I've read one finding on determinants of variation cross-sectional inflation expectations is that people who are about to buy something big (presumably strongly correlated with the y=1 observations here) are among those who actually have relatively accurate ideas about what inflation currently is, while those who are not about to buy something (presumably strongly correlated with the y=0 observations here) have much more dispersed inflation expectations. This would lead to the finding of the current paper. I guess one would describe this as an endogeneity problem, with y (buying intentions) causing x (inflation expectations).

    [I've read=Coibion, Gorodnichenko, and Kumar, 2014 working paper. Paper is on determinants of Firm's expectations of inflation; not people.]


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.