Is zero inflation really desirable?
To be more precise: Does inflation in real life have benefits that in some situations outweigh its social cost? E.g.: it works as a disincentive against holding money. The treasury derives income from printing new bills, which usually inflates prices.
The optimal level of inflation is very debated with unclear answers. There are many reasons, and a great answer would be very long. It should also distinguish between expected inflation and surprises.
I'm not going to do any of this, but giving you three reasons for a desirable positive level of inflation. This list is of course incomplete, also there are many reasons against too high inflation.
It is really unclear to economists why it is happening, but nominal wages seem to be downward rigid. It appears to be a behavioral thing (and might not be true at all, see Barro (1977)), but it appears that in crises, once presented with the choice of more firings or reduction of wages, most firms/workers decide to not cut wages but rather respond to the slump with increased separations.
To the extent that we believe this lack of reduction in wages is suboptimal, the central bank can enforce a reduction in real wages through increased inflation rates, and thus preventing separations to some extent.
This argument is based on Keynesian theory. Keynes claimed that the marginal propensity to consume out of income is smaller for rich households (and indeed, we do find this in the data to some extent). Unexpected inflation is similar to redistribution from creditors to borrowers, as long as debt contracts are not indexed to inflation.
To the extent that poor people consume more out of this unexpected wealth shock and than rich people decrease their consumption, this redistribution will lead to an increase in aggregate consumption.
This argument was most prominently brought forward by Krugman around 2008. In recessions, you want to be able to decrease nominal interest rates in order to punish households and firms for "holding cash" and incentivize them to spend it instead. If you start with low (say, 2-3%) nominal interest rates during normal times, you don't have a lot of space to cut back nominal interest rates during the crises.
If, instead, you would have higher nominal interest rates (Krugman argued for around 8%) and relatedly higher inflation during normal times, you could easier cut back nominal interest rates during busts and stimulate the economy.
Money is produced at zero social cost but held at a positive private cost (because to hold money you must forgo holding other assets). Therefore there is a positive externality from holding money, which means that the usual arguments for subsidizing it apply. This makes the optimal rate of inflation negative (because inflation is a tax on holding money). In fact if you add a few details to the above argument, it's easy to see that the optimal rate of inflation is minus the real rate of interest.
On the other hand, funding the government is likely to require departing from an optimal tax system in which everything is taxed (or subsidized) according to the externalities it creates. Presumably the optimal departure entails higher taxes on most things, including money holding, which means that the optimal inflation rate should be somewhere north of $-r$. Whether it's north of zero is less clear.
In the ECB's press conference of October 22, Draghi and his vice-president responded to the question of why the ECB is fighting low inflation so fiercely. Their answer provides a nice overview of why a low level of inflation is desired, and therefore should answer your question.
The following were their main arguments:
low deflation is to be avoided because it increases the real value of debt and leads to costly debt servicing;
measurement of inflation tends to be exaggerated by most price indices, therefore, targetting zero-inflation would imply deflation. This justifies why the inflation target should be above zero;
with low or negative inflation, real interest rates may be too high to reach full employment;
generally, central banks want to avoid deflation. The vice-president refers at this point to "real deflation" in contrast to "having negative inflation for a few months". This situation of "real deflation" should certainly be avoided, first, because of nominal rigidity in wages, and second, because it will lead to postponing consumption.
The transcript of the press conference can be found here: https://www.ecb.europa.eu/press/pressconf/2015/html/is151022.en.html
A little inflation is better than zero inflation (or deflation), because wages are "sticky" - they don't usually get adjusted downward.
If a company is facing a year that is a bit worse than the previous year due to small fluctuations in demand, wages should actually be decreased. Companies typically don't cut wages though, so they might even choose to let go of some of their staff. Now, if there is 2% inflation and wages remain flat in nominal terms (nobody is getting a raise), wages will be down 2% in real terms. So the company is better able to cope with the slump in demand (its real costs are lower) and it might not have to let staff go.
Moderate inflation also makes it possible to have negative real interest rates. This gives central banks more flexibility to stimulate the economy.
Real interest rate = Nominal interest rate - Inflation
If inflation is zero and the nominal interest rate is zero, the real interest rate will be zero and the central bank won't be able to make interest rates go any lower (assuming nominal interest rates can't be negative). However, if inflation is say 2%, if the central bank sets the nominal interest rate to zero, the real interest rate will be -2%, giving a significant stimulus to the economy.
In general, moderate inflation also helps keep the economy going. If there is inflation and consumers or companies sit on cash, their holdings are being devalued in real terms. This gives them an incentive to spend and invest instead of hoarding cash.
To decide whether zero inflation is desirable or not we should first qualify the question. Inflation is not a universal constant neither is the economic universe immutable so as to admit some particular fixed inflation level. Furthermore, inflation has different effects on different groups of agents; these depend on the composition of their wealth (debt or equity) and the nature of their incomes. Additionally, a second order issue might arise with the kind of time preferences that different agents employ when discounting the future.
Since late 2007-early 2008, the world economy has been perturbed severely by what has come to be known as the great recession. One major side-effect of this has been the systematic difficulty with which a number of Central Banks struggled to achieve the objectives of their mandates ie financial stability with subdued inflation.
In the past, the 'problem' with inflation had been its tendency to spiral uncontrollably upward. In this context, most recall instances before the Great Depression in the US, the Weimar republic's inflation sprout before the Great War, the stagflation of the late 70's-early 80's, Zimbabwe in 2008 etc.
This time around, the 'problem' is of the reverse kind. This time around, it is deflation (or perhaps, more appropriately, disinflation) that is the issue. The 'poster child' of this malaise is Japan and its two decades of deflation.
One characteristic of the present configuration that is up to par with the historic experience (to a greater rather than to a lesser degree) is the presence of debt build-up. Just like in the Great Depression and before WWII in Germany, debt in today's economies is by all accounts too high (debt is 'too high' not because of its absolute volume-which is admittedly high too-but due to the systematically subdued ability to service it).
Another characteristic of the current state of affairs is the subdued returns to labor income ('subdued' when compared to productivity). This is mostly related to the US economy and for a period of roughly fifty years since the early 70's but other industrialized countries show similar patterns for about the same period. This is the outcome of a number of effects such as financialization and innovations related to communications (internet) and transportation (containerization) that increased capital mobility globally. To some, the present situation is the result of the success of what has been called the 'neoliberal agenda' (related to the 'Washington consensus').
A third characteristic of the modern global economy is what has been called by B. Bernanke the savings glut hypothesis. The former Governor of the Federal Reserve, made a case of why interest rates are subdued in the US (and the world). As the argument goes, there is an oversupply of savings that are looking for productive investments but cannot be accommodated by the current state of affairs. Hence, real interest rates (a measure of opportunity cost for real activities) will be low for as long as the conditions responsible for the oversupply, persist.
Perhaps the most interesting approach of the current configuration is one put forward by Brown's Watson Fellow M. Blyth. Blyth talks about economic regimes of changing institutions (a-la Aglietta). Juxtaposing 'the 70's' with the present, he finds a shift in institutional set-up, from an early one aligned with debtors' interests, to the later one aligned with creditors' interests. Regimes change when the underlying institutions become fetters rather than facilitators. Our present day experience is-according to Blyth-a (potential) turning point.
The preceding exposition was an attempt to frame the current configuration in a short and succinct way. As it is claimed in the introduction, inflation is dependent on the particulars of each historic (economic) epoch. During the present 'regime', low inflation is a strain rather than a strong point in the system. It makes debt repayment difficult. In turn debt overhang subdues growth by dampening profitability and allows for the continued trend of productivity-compensation gap. Low demand acts as a constraint on investment profitability prospects which in turn constraints growth prospects even more. This vicious cycle is but one example of why zero inflation is not a desirable level of inflation, at the current juncture.
No one appears to have noticed that Gregory Mankiw's Principles of Economics answers this question on p 778-781.
36-4 Should the Central Bank Aim for Zero Inflation?
Let me know if you need me to quote all 4 pages here.
Related to your question, most of the CBs target an inflation rate near 2%. Laurence Ball argues in his paper that targeting 4% could be better and his arguments are solid in my opinion. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2468019
It became in a sense a reserve engineering for your answer but I think the paper will be beneficial to show the relationship between inflation and policymaker's tools and to see undesirability of zero inflation.
