Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday 28 April 2013

Why Inflation is not falling

There has been considerable interest in the recent IMF study that found that the responsiveness of inflation to the output gap (or equivalent measure) falls at low levels of inflation. But if the econometrics is right (and Nick Rowe has some cautionary tales here), what is the explanation for this? I start with two standard stories, but then suggest other possibilities that are specific to current financial conditions.

One standard explanation which the paper itself gives is based on the menu cost model of price inertia. The idea is that firms do not change their prices that often because there are costs to making any change (which economists call menu costs, perhaps betraying how often they spend in restaurants rather than buying food in supermarkets), and that often this cost might be higher than any benefit to profits in making a change. If you derive the aggregate relationship between inflation and the output gap from a model of this kind, the coefficient on the output gap will depend on how frequently prices are changed. So if price changes become more infrequent at low levels of inflation, the sensitivity of inflation to the output gap will fall.

Another quite plausible story which has solid empirical backing is that workers particularly resist nominal wage cuts. That actually implies an asymmetry in response rather than a general reduction in sensitivity (if the output gap was positive workers would happily see wages rise), but it will affect the average response in an econometric study that does not allow for asymmetry, and in current circumstances it is an entirely appropriate story.

You might think that enough, but I have a UK-centric reason for wanting more. In the UK, the ‘wages’ Phillips curve does not seem to be showing any reduction in sensitivity - indeed perhaps the opposite (although any additional sensitivity seems to predate the recession). That does not mean workers are not resisting nominal wage cuts, but the overall impact of this has either been small, or has been offset by something else.

The story I want to tell involves firms’ pricing behaviour, and the role of more risk averse banks. Suppose a firm sees demand for its output fall. Its profits are lower, but it calculates that it can reduce that decline in profits by cutting its price, if that price cut increases demand. There are two risks involved in doing this. First, the price cut might raise demand by much less than expected, with the consequence that profits fall further still. Once the firm realises this it can always put prices back up again, but in the short run profits will decline. Second, it may take time for the price cut to feed through into higher demand: those buying competing products may not immediately realise that they should switch. So although profits might rise eventually, they could fall in the short run.

So in both cases, there is a risk that profits in the short run might suffer as a result of the price cut. In normal times firms would be prepared to take those risks, either because the risks are symmetric (maybe demand will increase by more than expected), or because they represent an investment with a positive eventual payoff (as customers switch products). Critically, even if the short run might actually bring losses rather than profits, the firm’s bank will cover the losses because it is taking a long term view.

However, since the financial crisis, the firm may have noticed that the behaviour of its bank has changed. It refused the business down the road any credit, even though by all accounts its difficulties were clearly temporary. Although the firm would like to cut prices in the expectation that this will eventually raise profits, if the price cutting idea does not work out and the bank plays tough that could mean bankruptcy.

The idea is that the aftermath of the financial crisis, by raising the risk of bankruptcy associated with short term losses, has lead to greater price rigidity. In addition, there are two related effects that could actually lead to higher inflation in the short run. First, the firm does not like the fact that it can no longer depend on the bank to cover any short term losses. Who knows what might happen. So although a price increase might reduce profits if sustained (as customers gradually switch), in the short run profits will rise, and that allows the firm to pay off those debts which would otherwise keep its owners awake at night. This is the firm as a precautionary saver. Second, firms might be keeping prices low not because of existing competition, but because of the threat that a new start-up might emerge and steal some of its business. The one silver lining of the financial crisis for existing firms is that new start-ups are much less likely to get any money from the bank, so this diminished threat of new entry allows the firm to safely increase its profit margins.

I have absolutely no evidence that any of this has been happening, or indeed whether these ideas stand up to serious analysis. I don’t know of any papers that have explored the impact of financial frictions of this kind on prices, but that may well be my fault, so please point me to any you know. If there is anything in these ideas, then they caution against interpreting any current rigidity in inflation as evidence against demand deficiency.  


  1. I don't know about the UK, but in the U.S., corporate profits are at record levels and well above historical norms. Why should firms cut prices? For an economist writing down a single aggregate production function which is extremely simple, it is clear whether or not prices should be cut. But for the individual firm, which doesn't even know its production function because such a beast doesn't exist, how would the firm know that it should cut its prices and why do you think firms should do so?

    Now if profits were excessively low, then we could argue that firms should be cutting prices. With high profits, its harder to understand.

    Perhaps the question you should be asking is why are there not enough new entrants to take away excess profits from firms. Why is there not a boom in investment as a result of such high profit rates. Here, credit constraints, as well as natural barriers to entry (e.g. economies of scale) can well play a role.

    1. Still, they are not cutting prices. I see what you are saying and I think it is a good point, however I doubt that it stands as completely true.
      Margins are so high not because pricing is adequate, but because costs are so low. There is no investment, wages are down the rack, ofc profits are sky-high.

  2. I've looked at the data. From 2007-2012, 10 year gilt yields have fallen from 5% to about 2%. Manufacturing PNC net rates of return have fallen from 10.0% to 4.7%, a swing of -2.3% over 10 year bonds. Service PNC net rates of return have increased from 16.0% to 16.2%, a swing of +3.2% over 10 years. Clearly something different is at play with manufacturing versus services, but services at least appear to be earning excess profits given the low yield environment. Perhaps there was an increase in the general risk premium as a result of the crisis. Data for rates of return is from tables 6 and 7 from the ONS report "Annual Rates of Return for Private Non-financial Corporations".

  3. Interesting theory. I haven't heard it before (which might not mean much). I vaguely remember reading some paper(s) on macroeconomics of limit pricing in which markups were higher in recessions, but I don't think they brought financial frictions into play in the model.

    My hunch is that empirically it doesn't work. Canada seems to have had a similar increase in the flatness of the Phillips Curve, but the financial sector stayed in relatively good shape.

    There is something weird about the UK. But it's more the drop in productivity. GDP dropped much more than employment. I think that fact has gotta be related to the puzzle of why inflation stayed high in the recession.

    1. Nick, try using the equation... (unit labor costs = labor share * price level).
      Here is a graph at FRED that I use for the US.

      The equation holds steady for the data in the chart. Inflation is connected to labor costs.
      If inflation stays high during a recession, this means that unit labor costs rose, or that labor share fell... both in relative terms. In the US, unit labor costs fell after the recession. Labor share fell but inflation also fell some too.

  4. You have to think even deeper... Look at the equation for unit labor costs...
    unit labor costs = labor share of income * price level

    Unit labor costs are holding steady. Labor share is holding steady. Thus price level is holding steady.

    Utilization rates of labor and capital are rising in the US and eventually unit labor costs will have increase. Usually this is done in an economic contraction. As unit labor costs increase, we can either have inflation or higher labor share.
    The real problem is that unit labor costs fell after the recession and have not rebounded, like they should. This has kept the real values of labor share and inflation low. As utilization rates of labor and capital increase, relative unit labor costs decrease, which has been matched by declines in the labor share and some in inflation.

    The true secret to raising inflation is to slowly raise the labor share rate in such a way that inflation can keep up. If inflation drags behind too much, aggregate supply can go into a contraction. The secret is to raise labor share steadily and slowly.
    I use the AS-ED model for this analysis...

  5. Taking your story a bit further:
    Firms might anticipate that if they cut their prices, rivals will quickly follow suit. This is even more the case given that all firms are under the same short-run financial pressure so they would be more responsive to price cuts than usual. Hence, firms do not cut prices since they realise it would not bring in additional business, not even in the long-run, but might simply trigger a short-term price war with close rivals. Moreover, with this price-war threat, there can be a coordination failure, whereby firms are stuck in a bad equilibrium with low prices, but none of them want to unilaterally increase their prices because they are afraid the rivals will not follow suit this time in order to keep price low and steal business from the moving firm. That is to say, you end up with prices being sticky upwards whilst they can fall precipituously downwards in case of temporary price war to keep market share. In a nutshell, here you have what looks like a cartel but without the anticompetitive intent, it is just firms trying to hold on their market share and keen not to trigger a foolish price war. Quite theoretical I know, but I find it more plausible, or at least as plausible than your story which implies an element of market power, i.e., in the absence of the discipline provided by the threat of new entrants.

  6. There is probably a huge difference between MNCs and SMEs.
    MNCs have good profits (look to flatten at the momnet so see what happen in the future). And in the uS at least are sitting on a lot of cash.
    SMEs in general are the most difficult borrowers to judge for banks. So from that angle it is logical that if doubts are there (economy of the country in general, may be even currency, sector etc) they will cut it back.

    Seen the fact the MNCs are continuing outsourcing and faster than real growth plus productivity increase (at least in the US) future jobs have to come for a big part from SMEs. There is however no clearly directed policy for these. Eg they pay overall considerably more taxes than MNCs.

    Banks are simply pretty dysfunctional institutions. They like to improve their BSs however unlike a sector where most of the companies (effectively) went bust they do it mainly by buying cheap (via CB and selling relatively high). One would expect considerable costcutting in eg staff and salary levels in general but we hardly see those happening.
    In that picture there is also in a lot of countries not much desire to increase lending to the SME sector. CB behaviour is making a carry trade in especially the South look attrective (at least if you forget about the extra risk).

    Unlike before in times of crisis worldmarket prices for a lot of products did not drop that massively. Oil is eg still about 100 USD. Simply because EMs are doing relatively fine. Combined with the fact that since the crisis eg Chines wages have almost doubled which means a lot of the stuff simply got a higher costprice. Made even worse if you take the currency into consideration as well.
    Probably inflation in the pipeline has deferred a drop import prices have shown considerably higher increases in a lot of countries than their inflation.

    Retail is pretty inefficient in Europe compared to the US. Which might also not push a fast adjustment.

  7. Interesting, but I don't think that banks "playing tough" has much to do with it, I think the same story can be told of firms concerned with cash flow. The line of credit may have never been available, still bankruptcy concerns increase when there is an exogenous shock in demand. I've seen this pattern with my firm, which ironically makes banking software. For the last few years banks have been desperate to make loans, be because the same exogenous demand shock has dried up their asset base (that too is anecdotal, not data based, unfortunately)

  8. our universe of data of financial crisis is not confined to this one. What happened to inflation after the great crisis of 1929-31 and is this consistent with your hypothesis ?

  9. Interesting post. Reminded me of Broadbent from last year, different stories, but of the same genre perhaps!

  10. Prof. Simon’s claim that workers’ resist nominal wage cuts understates the point, I think. In the UK over the last three years or so employees have managed to get 2% or so annual pay increases despite their productivity falling. And same applies to entrepreneurs and senior executives, no doubt. And Oxford economics Profs? :-)

  11. I'm not an economist, but just thinking about the phenomenon of price rigidities at a time of weak demand leads me to speculate that when demand is weak, elasticity decreases. One major difference between now and earlier periods, such as the 1930s, is that due to physical, time, and other constraints, people nowadays are less likely to produce their own goods. Take clothing for example, in past times people were more likely to make their own clothes if they didn't have much money. Food is another example--most people nowadays don't have the option to grow their own food. So this would suggest that when demand is down, people are buying more or less what they need, and cannot make do with much less. This may be why, as evidence suggests, people spending out of their savings. If prices were to decrease, people may not necessarily buy more of the products whose prices decreased; they may save the extra income instead.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.