Tuesday, November 20, 2012

Let's have a little chat about inflation...

Inflation is one of those things that almost nobody who isn't an economist seems to understand (though that doesn't mean all economists understand it either!). First of all, there is the fact that most people don't even seem to know what inflation is. Some people seem to equate the word "inflation" with "a decrease in [my] real wages" - if things seem harder to afford, then it must be "inflation". Others seem to think that if stock prices go up, then that is "inflation". Still others use the word "inflation" to describe specific price changes - for example, if oil gets more expensive, people call that "oil price inflation". And the other day, I had a friend tell me: "[It's] simple. Low [interest] rates mean the price of money is low. That's a form of deflation."

(Note that that last statement is the exact opposite of correct. Inflation is a decrease in the value of money. Deflation is an increase in the value of money.)

Or the other day, someone on Twitter asked me: "How is it possible for inflation to help debtors when wages are going down? If wages are going down, doesn't inflation just make it harder for people to pay off their debts?"

The answer is no. Here's why. Suppose you make $50,000 a year and you have $50,000 in debt. Your debt-to-income ratio is 1. Also, just for convenience, let's say the general price level starts out at "1".

Situation A: -50% real wage growth, 100% inflation.
In this case, the new price level is 2. Your new real wage is $25,000. Your new nominal wage is $25,000 x 2 = $50,000. Your debt is still $50,000. Your debt/income ratio is still 1.

Situation B: -50% real wage growth, 0% inflation.
In this case, the new price level is 1. Your new real wage is $25,000. Your new nominal wage is $25,000 x 1 = $25,000. Your debt is still $50,000. Your debt/income ratio is now 2.

Situation C: -50% real wage growth, 50% deflation.
In this case, the new price level is 1/2. Your new real wage is $25,000. Your new nominal wage is $25,000 x 1/2 = $12,500. Your debt is still $50,000. Your debt/income ratio is now 4.

So as you can see, even if your real wage is going down by 50%, it's better to have inflation than no inflation if you are a net debtor. Inflation erodes the value of your debt no matter what is happening to your real wages.

So what's going on here? Why do so many people misunderstand inflation? Maybe it's a form of "Stockholm Syndrome". Inflation-hawkish economists have been bellowing, so loudly and so vehemently, that inflation is Satan - this goes back at least a hundred years - that non-economists just can't help believing it. People end up trying to think up reasons why inflation must be bad after all. When you offer them freedom - when you tell them that sometimes inflation can erode debt, relieve balance sheet recessions, and help stimulate the economy - they don't want to take it. , and they come up with more brilliant ways to identify with their captors. Or something like that.

I don't know. I feel like that's kind of a weak theory. What's really going on here? Why don't most non-economists seem to get what inflation is or what it does?


  1. Anonymous12:28 AM

    Which readers is this post meant for? Those who understand inflation? Or those who don't?

    If it's for people who don't understand inflation, you might want to explain the connection between inflation and wages (and how you arrived at your nominal wages in your examples).

    As an answer to your blog post in general: most people learn inflation as "it takes more money to buy the same thing." Learning the difference between "nominal" and "real" value, or learning how inflation works on a macroeconomic level, takes more time, and much of it is counter-intuitive. So, judging by the simple definition they learn, of course people think inflation is bad. (Are you concern-trolling, Noah? I thought it was widely known that the "takes more money to buy the same thing" misconception dominated non-economist views of inflation. Are you trying to present inflation being good as matter-of-fact to troll the inflation hawks? If so, I apologize for this comment.)

    1. If it's for people who don't understand inflation, you might want to explain the connection between inflation and wages (and how you arrived at your nominal wages in your examples).

      Done. Check the examples to see how it works.

  2. I think it's simple. Inflation means "the price of things is going up". "Hey, *I* buy things. I don't want the price of the things I buy to go up." People think about the price of the things they buy every day. They get a chance to negotiate their wages way less frequently, so they don't think of their own income as one of the prices that go up.

    1. Lawrence's comment reminds me of this post from Yglesias, on how perceptions of inflation has changed since the US economy has restructured: http://www.slate.com/blogs/moneybox/2012/05/24/everyone_hates_inflation_because_they_don_t_understand_what_it_is.html

    2. I usually say "It's not general inflation unless your wages go up. If you're wages aren't going up, it's just price-gouging."

      This does raise a very important issue. We could print ten trillion dollars, and *if they were all given to banks or big corporate CEOs*, it wouldn't cause inflation and it wouldn't help the economy one bit. Only if the printed money is given to *poor and middle class people* will it actually help.

    3. MMT! :)
      Fed buys all outstanding US debt (mostly from the Wealthy and Banks), holds the debt, pays the profit on the debt too... the Treasury! :)

      How does that play into the classic money supply inflation model?

      The classic simple example of inflation is this: The Fed (or the banking system, giant Rothschild robots, whatever) doubles the money supply. So money becomes only half as valuable as before. Therefore, the price of everything doubles. So:


  3. What I think is happening here is that the constant cries of "inflation!" are being confirmed in the eyes of the "captives" in two ways. One, there was the 1970s inflation that really was damaging. Two, a reduction in real wage growth makes it seem like something similar is going on.

    What's the price that most people are familiar with? The price of gasoline, of course. Look at any Facebook feed and you'll see an uncle here or there posting pictures of gasoline station signs from yesteryear.

    It's a pretty convincing story to someone who isn't an economist or took their last economics course 30 years ago. They don't feel as wealthy, they feel the gap closing between the US and other nations, they see the most salient price in the American imagination only going up, up, up...

    Half of Michele Bachmann's campaign was squealing about $2.00 gasoline. A lot of people were excited about that.

  4. Anonymous12:57 AM

    Noah, I feel like you have left out situations D, E, F, G, H etc. And you have mysteriously set up the problem as though the only relevant economic factors affecting well-being are income and debt.

    What about consumption expenditures and real purchasing power?

    What about savings?

    Use some more realistic real wage and inflation rate numbers, then add in the effects on what you are able to buy and what happens to your savings, and you get very different outcomes. Benefits in the area of debt pay-down can easily be offset by the impact on purchasing power and savings. If prices go up and my nominal income stays flat, as in your situation A, then by debt/income ratio is unaffected - but all the stuff I buy with that nominal income costs more. Either I have to buy less or buy the same amount and save less.

    That said, I'm all for engineering an increase in the price level if it is done in such a way as to be accompanied by a corresponding increase in nominal wages. A helicopter drop into worker bank accounts across the country would create inflation, but also boost those workers' nominal incomes. Could be worth it.

    But I know for a fact that the biggest boosters of central bank induced inflation - to the extent they even have a plausible non hand-wavy, non-expectations magic transmission mechanism in mind - have explicitly said they want to depress real wages further. So I don't trust them.

    Look, I know you are a really young guy, but you have enough training and have no business being thick about impacts on people who are not so young. Do you have any idea what it means for a person who is five or ten years from retirement to see half of the retirement assets disappear in a puff of Wall Street smoke during a catastrophic financial meltdown? Do you really not get it about why people are not comfortable with some blunt tool monetarist inflation scheme that will further dissipate the value of the part of those assets that weren't vaporized in 2008?

    Also, if we could get a sort of pure, uniform, absolutely steady inflation at some high rate, uniform in that wages and other prices are all effected equally, then the inflation might not be any big deal. Every contract would just have the inflation premium priced in and things would go on as usual. But the other reason people worry about inflation is that, silly monetarist textbook models of omnipotent central bankers to the contrary, real accelerations in inflation tend to be volatile and unpredictable. When people can't predict and price the value of a contract, risk-aversion sets in and they stop making them.

    1. Dan, Noah included most of that. Savings is the opposite of debt. (So, YES, someone with savings and no debt is hurt by inflation.) Real purchasing power is exactly what inflation is measuring.

      The final issue you raise is interesting, however:
      "When people can't predict and price the value of a contract, risk-aversion sets in and they stop making them."

      This is the primary argument for near-stable money. Not for truly stable money, but for "near enough to stable".

      Note that it's also an argument against DEFLATION. Deflation also prevents proper pricing of the value of a contract.

    2. "A helicopter drop into worker bank accounts across the country would create inflation, but also boost those workers' nominal incomes. Could be worth it."

      Absolutely worth it. I suggest $200 a week, mailed directly to every citizen of the United States, for the indefinite future. When poverty in the US is eliminated or full employment is achieved, we can cut back.

      I expect howls at this suggestion, but I think it would probably work.

    3. Anonymous7:51 AM

      neroden, we should add that with so much unemployment, and with so much unused capacity not as yet destroyed, it's entirely possible that a helicopter drop will not being inflationary but only usefully boost demand.

      I see Noah did draw his conclusion only about net debtors, but he cherry picked his examples to suggest that most people benefit from inflation, and left out consumption purchasing power. Some of the less well off in our society consume a very high proportion of their total income on necessaries. One way to think of those necessary expenditures is as a kind of monthly debt payment. But in this case it is not debt at a fixed nominal rate, but an adjustable rate that ticks up monthly in exact proportion to the rate of inflation. If their nominal incomes stagnate as that adjustable rate moves up every month, then the net negative effect might offset the positive effect on their debt bill.

      Agree about deflation.

    4. Don't most of the people who are less well-off live on inflation-adjusted incomes? Social security, food stamps, etc are indexed to inflation.

    5. Anonymous9:32 AM

      Consumption purchasing power is measured by real wages. In all three examples, you become 50% poorer, whatever happens to the price level, and the only relevant question is whether your debt becomes more of a burden (or how much your savings are eroded, if you're a net nominal creditor).

    6. Anonymous9:51 AM

      Anonymous, you can't run a sound dominance argument if you don't consider all of the alternatives, and if for each alternative you only consider a partial list of the relevant outcomes.

    7. Anonymous9:58 AM

      JSR, I'm not speaking about people who live on transfer payments. I'm speaking of working people who don't make a lot of money.

    8. "But I know for a fact that the biggest boosters of central bank induced inflation - to the extent they even have a plausible non hand-wavy, non-expectations magic transmission mechanism in mind - have explicitly said they want to depress real wages further. So I don't trust them."

      Actually the biggest boosters of expansionary central bank policy advocate dropping inflation targeting altogether and have stated that they explicitly want to raise nominal incomes. And since real wages tend to be procyclical, a more expansionary monetary is much more likely to raise real wages than decrease them. However if you have something to support this extraordinary claim please feel free to provide it.

    9. Anonymous2:10 PM

      Yglesias last year:


      Monetary expansion to lower real wages but improve debt/income ratios.

    10. Anonymous2:11 PM

      I like Noah's May post better:


    11. Actually I think working people with low incomes are the most likely to benefit from an inflationary policy at the moment. They are likely to have a larger debt/income ratio than average; they are likely to have lower savings than the average; since they are not living on a fixed income, their earnings are likely to go up in nominal terms; if inflationary policy improves the job market, their earnings may outpace inflation.

    12. Depressing, however Yglesias is not an economist. He's taking a story that's mainly about flexible exchange rates and interpreting it in terms of real wages. Expansionary monetary policy only raises prices if it increases aggregate demand, and if aggregate demand increases so likely will real wages.

  5. Noah,

    Situation is rather different over here, in Israel. The largest debt any one person takes is of course mortgage, which is almost always CPI-linked. Wages, however, are only partly linked the the CPI, with the majority of wage earners receiving no compensation for inflation (In the long run wages SHOULD rise, but that long run sure as hell takes time to arrive).

    So no, when inflation rises, real wages get cut, while real debt remains the same (nominal debt rises). I assume from your example that this isn't the case in the US, but for shellshocked Israel (hyper-inflation in the 80s...) no bank will lend you anything for over 10 years without some sort of inflation-compensation mechanism.

    1. Well, that's nasty. Variable-rate debt is considered exceptionally dangerous in finance; forcing it on everyone is pretty nasty. Perhaps the best thing to do is to liquidate the existing banks and change the law.

    2. Anonymous2:13 AM

      This is strange to me given what very little I know about Israel's central banking. IIRC, they seem to have had a de facto NGDP targeting regime for about a decade now. Sure takes a while to build up that credibility!

    3. neroden,
      That is why many Israelies are starting to look for fixed-interest financing. The only problem is that with the booming prices of real estate in Israel, many people won't be able to afford the initial payments of the (higher) fixed-interest loan.

  6. It's absolutely political.

    Inflation is good for debtors and bad for creditors. (Debtors == more debt than savings; creditors == more savings than debt).

    Creditors are, by definition, richer than debtors.

    Rich people can pay economists to bellow. Debtors can very rarely afford to do this.

    So, rich people pay economists to bellow about the evils of inflation, and they have done so for over 100 years; while the poor people to not pay anyone to bellow about the evils of deflation.

    Result: people think inflation is bad and know nothing about deflation.

    1. Second comment: the SPREAD.

      When your savings account pays 1% interest, but your mortgage charges 8% interest, the spread is 7%.

      Spreads are a transfer of money to bankers from the entire rest of the population. High spreads constitute extraction of wealth out of the entire rest of the economy, and transfer of that wealth to the banking sector (usually to the CEOs and other execs).

      Because the banking sector is very close to unproductive, acting only as a middleman, high spreads are inherently bad for the economy. We've watched this happen -- the "financialization" of the economy, it's called -- as a larger and larger portion of the national wealth is extracted by bankers.

      (There does need to be a small spread to pay for the actual services provided by banks: accounting, legal work, custody of bills and coins, etc. Many banks have been avoiding actually doing THIS work properly, lately.)

      Government action to reduce spreads would be extremely valuable. The old usury laws had the effect of reducing spreads, but the interest rates set in those laws weren't adjusted as inflation changed, which created problems in the '70s.

    2. Anonymous10:23 AM

      Well, it's not straightforward that creditors are richer than debtors. In many developed countries (and the USA may be an exception here) it's hard to get a loan when your income is too low. The poorest ones in Europe typically rent the place they live in and have few savings on a bank account. Moreover they are likely to work in a low productive industry where wages are unlikely to increase with inflation. Inflation may be harmful for them as they have less opportunities to protect their wealth and less bargaining power to renegotiate their wage.

      Redistributive effects of inflation are particularly hard to estimate and they may vary from one country to another.

    3. neroden@gmail12:38 AM

      By definition, creditors are richer than debtors. This is the definition of "richer". People who have savings in a bank account are richer than people with a large debt (perhaps student loans) and no savings.

  7. Noah, it seems to me that you have made a pretty good case for never using the word inflation again or perhaps even the phrase "price level."

    Both of too general to be meaningful descriptors of anything.

    Does any other science use such a broad term that no meaningful information is conveyed?

    Cannot a good argument be made that a great deal of the problem with Macro is that it is so unscientific that it has never developed good tests regarding the health of the economy so it speaks only in broad generalities and averages that are meaningless?

    Examples, in a healthy economy, what should be the debt/income ratios for the various sectors of the economy (and how, scientifically, should those sectors be identified?)

    In a healthy economy, how stable should prices be for basic commodities?

    In a healthy economy, how much substitution of one commodity for another is ok? (natural gas for oil)

    It seems that polling from the recent election offers a great insight. Some people got the election wrong because they choose to use the wrong data and then treat such in the wrong way (I leave it to Brad was to the plan, design, or intent). Others selected data, treated such in certain ways, and got at least an accurate prediction this time (luck v. good?).

    Instead of "inflation" or "deflation" where are the truly descriptive measures. The patient's potassium level is X---therefore---is a lot more informative than the observation that you look flush in the face.

    1. "Examples, in a healthy economy, what should be the debt/income ratios for the various sectors of the economy (and how, scientifically, should those sectors be identified?)

      "In a healthy economy, how stable should prices be for basic commodities?

      "In a healthy economy, how much substitution of one commodity for another is ok? (natural gas for oil)"

      Why would you think there are general, "scientific" answers to such questions? I think the answer in each case is "it depends."

    2. Anonymous12:58 PM

      To paraphrase Richard Feynman, if its not science its bullshit.

      Simply put, there is too much incentive caused bias in economics. Look at all the people pulling the rope for Romney on the hope that there were appointments to be had to Treasury or the Fed.

      Or, look at the all the people, starting with Hayek, who attack Keynes merely to draw attention to themselves. My favorite in that category is Williamson, whose first post was a vicious attack on Krugman, intended solely to draw attention. He hasn't stopped claiming that he can use a model that cannot predict bubbles but can predict inflation.

      I could go on, but I believe sufficient examples exist to make my point.

      I do not believe that there are general scientific answers, for I tend to think that Minsky was right and that all economics is a confidence game. If I were bright enough to come up with the answer, well I suppose there would be a Nobel prize in it for me, but everything I have read convinces me that at some point finance shifts from being investing to being gambling. I believe that a lot of the answers lie in the microeconomics of lending (recourse, non-recourse, equity kicker, etc.) and that we could find the best rules through trial and error experiments, but absolutely no one is willing to talk about doing anything big in economics.

      Look at Greece or Ireland. The countries are so small that for the last 5 years we could be using either or both to learn. Greece has fewer than 12 million people, yet no significant economist has ever proposed the equivalent of the LHC, which cost $9 to 10 billions.

      Why didn't economists seriously propose that the World offer Greece to pay the interest on its debt for 7 years in exchange for being an experiment in entrepreneurship, taxes, banking, etc.

      I could go on, but what's the use, the vision is so limited

    3. It would be unethical for economists to experiment with with people's livelihoods that way.

      Also, if you're going to insist on conducting experiments some sort of MMORPG would probably be cheaper and more robust since you can more easily track transactions.

  8. Noah, here it comes the discussion between price and value (not in marxist's terms). We have to discriminate money (i) as an instrument used to buy goods and services (the value of money) and (ii) as a good itself (the price of money). Note that I am deliberately not considering a third category: money as an asset.

    When your friend told you about the interest rate he was talking about a measure of the price of money, the second case. Although your relation between the value of money and inflation/deflation is logically correct, the context where it belongs, I mean, as your interpretation of your friend's statement deals with the first case.

    We could, thereafter, complete the argument: inflation is a decrease in the value of money (case 1) and an increase in the price of money (case 2). Deflation is an increase in the value of money and a decrease in the price of money.

    1. No. The interest rate is the price of waiting, not the price of money. The price of money is just the amount of goods you must trade to get some money, i.e., it IS the value of money.

    2. I think price and value are two distinctly subtle concepts, albeit interrelated. Price is a signal that measures the capacity of the economy to attend the demands of the society, i.e., it reflects real factors and it's determination comes from the balance between supply and demand. In this sense the interest rate is the variable which balances the supply of and demand for funds (The Loanable Funds Theory).

      As Friedman (1955) points out, the value is determined by the marginal utility, what implies the subjectiveness of the concept; on the other hand, the value in exchange (relative prices)is determined by the marginal rate of substitution between two goods.

      "The price of money is just the amount of goods you must trade to get some money." But if you don't have any good to trade, how can you get some money? You'll have to borrow it, and for this to be true the price of money should be something different.

      By the way, in microeconomic and psycological terms, I would argue that the interest rate is a measure of the opportunity cost faced by the owner of the money. And the individual who wants to borrow some money should assess whether the marginal utility provided by this amount of money more than offsets it's cost.

  9. Noah,
    Nice post. I think the vocal concern comes from retirees, whose incomes are not necessarily indexed, and who do not have significant debt.

    Macroeconomic aggregative effects are harder to understand and mostly the middle class does not give a f*k about it, since what really matters to people is that they are paying more for gas and food.

  10. I've been waiting for a post like this. Here's what I think the problem is. We always talk about sticky wages as if there is only a floor, but perhaps within the job you currently have there is also a ceiling. Just as the employee doesn't want to see his nominal wage decline the employer doesn't want to see the nominal wage incline. However, if a firm needs new employees, they are obviously going to need to pay the market price, and for new employees the wage will obviously go up. So in order to actually see the benefits of an increased nominal wage from inflation you have to switch jobs, something people don't really like to do. And when they do find a new job at a higher wage they aren't going to assume it was because of inflation that they are getting a higher wage, they are going to assume its because they are such a valuable worker. So I think people dislike inflation because it hurts stability.

  11. Anonymous9:55 AM

    I think the root of the problem is the 1970s. People's income stagnated while prices grew exponentially. It was a tremendous shock to a Western world that had grown highly accustomed to massive economic growth.

    The correlation of extremely high interest rates, stagnating wages, unemployment, and recession at that time created an inflation-phobia in the Boomer generation.

    (It also seems to have resulted in the near-deification of Reagan and Friedman when the hard times finally "ended".)

  12. You forgot to talk about inflation and saving....

    1. Does inflation affect saving?

    2. "Does inflation affect saving?"

      More accurate to say saving affects inflation. Pople don't spend until their desire for saving is sated, which means simply pumping dollars into an economy does not necessarily lead to wage/price increases. If the private sector's savings rate is high, then government spending must be high to satisfy it (as the central bank itself cannot inflate). Then and only then will addiional dollars be spent, and inflation will only occur once aggregate demand exceeds productive capacity.

    3. Inflation is a tax on savers.

    4. Only if you save your money at fixed nominal interest rates.

      If you save your money by putting it in the stock market or in commodities, then inflation is not really a tax on saving, is it?

    5. yes of course =) !!!

      but then one begs the question why people (risk averse passive depositors) don't do that!

    6. Ben Johannson6:12 PM

      They don't do it because putting one's money into stocks and commodities is not considered saving, but investment. Investment is a form of spending because you're buying an asset, whereas saving is income not spent.


    7. but then one begs the question why people (risk averse passive depositors) don't do that!

      Dunno...accepting pointless inflation risk by demanding fixed-rate nominal contracts seems risk-loving to yours truly, but what does yours truly know...

    8. neroden@gmail12:42 AM

      Noah, talk to investors in Enron. Fixed-rate nominal contracts are a hell of a lot safer than, you know, Enron!

      Remember that the marginal utility of money declines as you get more of it. Guaranteeing that you preserve your $1000 is *more valuable* than a 50% change of doubling it with a 50% chance of losing it.

      This simple fact -- the marginal utility of money declines as you get more of it -- is well-known in economics, yet I've never seen a single economic model which incorporates it. You can't predict any "rational" economic behavior without incorporating it. This goes to show, again, that most economics is bought-and-paid-for right-wing hackery.

  13. Anonymous10:52 AM

    Three quick questions.

    1. Just out of curiosity, does anyone know when inflation changed from being an increase in the money supply to an increase in some kind of price index? If my memory is correct, Locke, Hume, and Smith discussed it as the quantity of money.

    2. Do creditors matter at all in this analysis?

    3. You bring up an almost irrational fear of inflation amongst certain individuals. It seems to me many economists have had the same sort of fear of deflation. Is the fear of deflation at any cost just as irrational? If not, why not?

    1. Ben Wolf1:34 PM

      1) So far as I'm aware inflation has always been defined as a rise in prices. Classical economists considered expansion in supply of money as the direct and proportional cause of inflation, as neo-liberal economists do for the most part.

      2) Only insofar as inflation is better for them than widespread defaults to discharge debts.

      3) Deflation suppresses consumer spending and wages, particularly when the private sector is heavily leveraged. I'm not aware of any example of deflation leading to improved well-being for a society.

    2. Anonymous3:33 PM

      "2) Only insofar as inflation is better for them than widespread defaults to discharge debts."

      How do we know this?

      " I'm not aware of any example of deflation leading to improved well-being for a society."

      Are you aware of an example of inflation leading to an improved well-being for "society"?

    3. Ben Johannson6:08 PM

      1) it's better to get some of the principle back than none.

      2) You asked why deflation is considered bad. I answered.

    4. Anonymous9:34 AM

      1) Wouldn't the individual creditors and debtors determine the best way to get some money back by perhaps restructuring the loan? Therefore, some principle can be repaid and it does not require widespread effects that accompany inflation.

      2) To go back to an earlier point, can you point to an example of deflation making people worse off? I'm not looking for a correlation between contraction and deflation, but evidence of a causal relationship. There probably is some example, I'm just looking for it.

    5. neroden@gmail12:43 AM

      econpointofview: Look at every 19th century recession in the US for examples of deflation hurting people. Look at WWII and the Civil War for examples of inflation helping people. Really, this is trivial.

  14. Anonymous11:08 AM

    What is interesting, in my view, is that people ALWAYS tend to associate inflation with lower real wages, while the evidence (at least in the short run) is mixed (or even contrary in some developing countries with median rates). Sometimes this view is shared by many economists.
    So the public perception of inflation is in fact worst that it really is (compare to, for example, unemployment).
    As you say, Noah, this maybe the result of the satanization of the phenomenon by media economists, but we can add perhaps a psychological twist: all the people is exposed to inflation with every purchase, but the fear to unemployment is confronted just by some, and on specific moments.
    If there is a tradeoff between inflation and unemployment, what kind of policy must we follow? Taking care of the (wrong) feelings of everyone, or put first in the agenda the true suffering of the fragile ones?

  15. I (and Scott Sumner) have publicly advocated abolishing inflation (the concept, not the phenomenon) for the very reasons alluded to in this post.

  16. Anonymous12:44 PM

    It seems to me that Noah is committing a decision-theoretic fallacy here. Let me illustrate.

    Draw a 4 by 4 table representing the payoff matrix for a particular worker. At the top, put "inflation rate" and on the left put "nominal wage increase." Across the top margin, enter the four alternative inflation rates of -5%, 0%, 5% and 10%. And similarly, down the left margin enter the three alternative nominal wage increase rates of - 5%, 0%, 5% and 10%. Now assume the following starting point, in nominal terms:

    Annual nominal wage - $50,000
    Total debt - $25,000
    Annual consumption bill - $48,000

    Now for each of the 16 cells in the table, compute the new nominal wage and the new annual consumption bill. (Assume that the debt is fixed in nominal terms and so remains $25,000 for each outcome.) Now compute both the debt to income ratio and the consumption to income ratio for each of the 16 outcomes.

    Note that there are six outcomes in which real wages fall, those in the upper right hand corner above the diagonal. What Noah argues is that if real wages are falling - i.e. we have an outcome in the upper right-hand corner - then one does better to have inflation than no inflation. In other words, he is arguing that of those 6 outcomes, the five in which there is a positive rate of inflation are superior to the one in which inflation is 0%. But there are two things to say about this conclusion:


  17. Anonymous12:45 PM

    (cont. from above)

    1. As anybody who has considered these kinds of two-dimensional payoff matrix problems knows, one doesn't evaluate whether a particular outcome generated by the two independent variables is an optimal one with respect to one of the variables by comparing that outcome to other outcomes in the grid with the same value of some dependent variable - which is what you are doing when you compare one of the outcomes in the upper right corner with other outcomes in that corner. Instead, you consider what happens when you hold one of the variables fixed and vary the other one. or each outcome in this particular grid that corresponds to a positive inflation rate, the worker in question does better with the same nominal wage change and a lower inflation rate. Noah is treating inflation rate and real wage change as though they are two independent variables. But as a consideration of his three outcomes shows, these two variables are not independent, since the rate wage change is a function of the nominal wage change and the inflation rate.

    2. Noah has left out the consumption bill entirely. So perhaps he is giving a ceteris paribus argument assuming the real consumption bill stays constant. But it doesn't. It is also a function of the nominal wage change and the inflation rate. This complicates the evaluation of the outcomes since we are comparing changes in the annual real consumption bill with the change in the total real debt. But I hope we have at least enough here to show why Noah can't derive his conclusion.

    3. So in the case of individual workers, it should be clear that the evaluation of the impact of inflation is more complicated than Noah paints it and depends on their circumstances. If we instead consider this table as not the table for a single worker but the table for a representative worker and the who macro economy, then one issue is whether we can choose policies that move us in a diagonal direction with higher inflation and wages concurrently. Macroeconomists, I gather, tend to make this assumption as a matter of course since they treat a concurrent increase in all prices - including the price of labor - as part of their definition of "inflation". But here I think Noah needs to follow the Michael Moore lesson he stated before so eloquently in his previous post, and get out of the clouds of idealized macroeconomic aggregates to look at the details of the likely actual dynamics of realistic phenomena of which those macro models are only very much oversimplified idealizations.

    I have been reading posts like this for a couple of years now both here and other blogs, like Matt Yglesias's blog. There is a persistent line of argument according to which some kind of simplified textbook analysis yields the broad result that "inflation is good for debtors". Then the writers and commentators go on to say, "What is wrong with those older folks and their inflation fears? Must be some kind of irrational inflation trauma pathology!"

    But maybe some of you guys should consider that you need to do better and more realistic math and mathematical modeling, and that you can't explain the attitudes of the people who disagree with you solely by virtue of the trauma inflicted by memories from the 70's of Jimmy Carter's Win buttons?

    1. neroden@gmail12:45 AM

      I think the key point which should be made here is "It's not inflation if your wages don't go up, it's just price-gouging".

      So, y'know.

  18. In popular culture, inflation is when other people raise their prices, growth is when you raise your own price.

  19. Nice examples, even though it is worthwhile to point out that inflation does not generally errode incomes except for those on fixed income.

    By definition, inflation is a rise in the general level of prices. If we all raise our prices, then we will pay more for the goods and services produced by others but also receive more in exchange for the goods and services produced by us. It is important to remember that someone's income is someone else's expenditure. You cannot have inflation without a simultaneous increase in people's incomes.

  20. Another well informed post with a little dab of required history. Also, I think people would understand what happened in 2008 if they looked into history. 2008 was an EXACT repeat of 1896. L. Frank Baum immortalized the events of 1896 in the movie and book Wizard of Oz. If you want a little insight into recent events, rewatch the movie. It will make much more sense to you.

  21. Mr. Pinion, in my humble opinion you have totally overanalyzed the Twitter question and ended up speaking nonsense. The guy wasn't asking you about real wages, he was asking about actual nominal wages.

    So his point was something like this:

    "Hey Noah, I get how if all prices double, including my hourly nominal wage rate, then it becomes easier for me to pay off my fixed debt. But if I'm having trouble making ends meet and keeping up with my loan repayment plan, then my employer actually cuts my nominal paycheck, and then on top of that the price of food and health insurance goes up, how in the world does that make me better off?!"

    It doesn't. By interpreting your Twitter guy to be speaking of real wages, you completely dodged his simple question.

    Why is it that so many economists can't understand simple questions about inflation? Must be political.

    1. Bob,

      At first I thought you really had Noah here. Obviously the guy asking about inflation when wages are down was talking about nominal wages not real wages, so Noah’s example didn’t make sense. But as I thought on it a bit more I came to see that Noah was right.

      Suppose inflation is 10% and a guy gets a 5% pay cut. He might think “man, this inflation is going to make it even harder for me to pay my student loans after this pay cut.” And it’s true, if inflation were 0% AND the guy still only got a 5% pay cut, then he would be better off. But why assume that would be the case? The same forces that are pushing up prices should also push up wages such that, because of the inflation, the guy only gets a 5% pay cut instead of a 15% cut. So Noah was right to handle the case the way he did.

    2. Major_Freedom10:56 PM

      So....Mr. Smith did not actually answer the Twitter guy's question about how can inflation benefit someone whose wages are falling and owes debt.

      How is not answering the question asked "the right way to handle the case"?

      It is not even correct to assume that inflation ipso facto raises a given wage earner's wages the same rate it increases the prices of the goods he buys. Inflation does not work that way. Inflation raises some people's incomes and some prices before other incomes and prices. It is nonsensical to pretend that falling wages and rising prices cannot coexist. 1970s stagflation? Hello?

  22. Anonymous9:31 AM


    Situation A: RWG: -50% INF: 100%
    Situation B: RWG: -50% INF: 0%
    Situation C: RWG: -50% INF: -50%

    I am curious: Why 100% inflation in A, if all of the other nonzero adjustment values are 50% increments (presumably for the sake of simplicity)?

    What happens to the D/I ratio if we change that 100% to a 50%?

    Huh. Non-economists are such fools...

  23. Noah, you are being traduced over at Bob Murphy's. You might enjoy the thread -- you seem to have a refined sense of the absurd -- and we'd love to hear from you.

  24. I think inflation fears, or more precisely, a preference for a lower rate of inflation, have to do with estimated uncertainty of the effect of inflation in the context of risk aversion, leading to a status quo bias.

    Think about it this way. A hypothetical person in this thought experiment makes an inflation-adjusted $50,000 annually. A genie comes to him or her (and only him or her) and offers an extra 2% increase in prices for the next 3 years. In return, that person's real income will rise according to the average per-capita increase in real income in the counterfactual scenario in which aggregate demand rose fast enough to support that increased inflation in the short run. Would that person accept that offer? Would you accept that offer?

    You and I might say that we think demand is the real constraint on economic growth, and so we'd assess the average per-capita increase in real income at probably in excess of 6%. But it's really, really easy in a expected welfare model to make the person not accept the genie's offer with reasonable assumptions of uncertainty along the chain of effects and risk aversion. What happens, obviously, is that people prefer the certainty of their real incomes today rather than the probability distribution of outcomes inflation offers.

    This is why I think the argument that people shouldn't oppose demand-driven inflation because it shouldn't affect their real incomes as missing the point. The problem of inflation, as people see it from an expected-welfare perspective, isn't really the change in real income. It's the uncertainty cost. That will totally wash out any real-income effects on welfare.

    1. Eh, this is a pretty standard approach, so I get what you're saying, but I don't have a lot of confidence in it. Noah's claim that it has more to do with hype is a better explanation. A lot of the people most severely affected by inflation don't really understand concept - seriously, ask around.

      And if we're talking about uncertainty, we're talking about variance. Has inflation in recent history seen a lot of variance? Not really. Do forecasts of inflation show a lot of variance? Not to my knowledge. Our confidence bounds seem pretty tight; the Fed's pretty good at what it does.

    2. Major_Freedom11:01 PM

      I think uncertainty is washed out when someone is told they are going to get a pay cut, and they are fairly certain that price inflation is going to be positive that year. This person is guaranteed to lose.

    3. Ok, but that just brings us back to the start where no one is necessarily losing from inflation.

  25. If you haven't, you should read Iriving Fisher's 1928 classic "The Money Illusion" (yeah, Sumner did not invent that title)- it basically deals with the original question.

  26. Anonymous8:33 AM

    Surely part of the problem is that once economists start talking about policy, there's a big tendency for them to get very loose about their definitions of inflation and even looser about how it is measured.

    I had a great discussion some years back online with Mark Thoma on his blog about "core inflation measures" - he said "it's good, it weeds out volatility" - I said, "but at the expense of taking any kind of asset price inflation out of the measurement." Then we had a big crash related to the asset price boom...

    Point is not who was right, but that the whole concept is woolly once you get to measurement and praxis instead of just the theory...

  27. I have another question: when economists talk about interest rates which interest rates do they usually picture? The overnight rate?

    Also when we are talking macro do house prices get factored in? Because depending on the weight this might mean significant deflation around the housing crisis.