Wednesday, September 22, 2010

What You Need to Know About Inflation

Inspired by a thread on Reddit, I thought I’d explain inflation as far as it needs to be for most practical purposes.

Inflation is most commonly defined as (1) an increase in the supply of money or (2) an increase in the “general price level.” Deflation is obviously the opposite. Definition (2) is the most mainstream and popular while a minority of economists, primarily those from the Austrian school, will use definition (1).

But definition (2) is problematic. Prices rise and fall because of fluctuations in supply and demand for goods themselves and because of fluctuation in money supply. Referring to fluctuations in prices with the same word seems limiting and possibly confusing. If the price of televisions go down because productivity (and thus supply) increased, does it make any sense to use the scare word of “deflation.” Similarly, it doesn’t really make sense to use “inflation” to describe short term and generally innocuous increases in prices. To use “inflation” to mean “an increase in prices” is to focus on the symptom and to ignore the cause. Why then use “inflation” to describe “an increase in money supply?” The best reason is that money supply is a constant focal point. In the production of any good, there are a wide variety of component parts that affect the price of the final good, but as prices fluctuate, markets are usually able to keep them more or less stable, especially at the “general level.” Furthermore, the tendency (and entire point of markets) is for prices to decrease over time as productivity is improved. So the tendency of markets according to definition (2) is “deflation,” not as friendly word as you'd expect.

Conversely, the supply of money will always be positively correlated with price because goods are purchased in terms of the medium of exchange. And that medium of exchange has a single source: the government. Whereas prices on the market are determined by an infinite number of seemingly insignificant details, money supply is traceable right back to the centrally planned banking sector. However, just printing new money doesn’t necessarily cause anything to happen if it doesn’t circulate or isn’t used to backstop anything. Plus, in our current monetary regime, increasing the money supply can have negligible visible effects. For this reason, my preferred definition is “increases in prices from where they would otherwise be because of an increase in the money supply.” This definition incorporates both into one while explaining a cause and effect. And so long as the money is not buried in a ditch or sitting in a mint somewhere, it is affecting prices by either facilitating bidding on assets/goods or by propping up institutions as reserves.

The problem of inflation is not, however, solely an issue of increasing the money supply. In fact, this is rather mundane by itself. The problem is with the way money supply is increased.

If, for example, every year the Federal Reserve increased the money supply by 5% and (assuming no one had $0, interest rates and prices were allowed to increase accordingly, and the monetary increase is expected by all participants in the economy) everyone’s cash holdings were increased by 5%, we’d expect prices and wages to all increase by 5% as well. Increasing the money supply proportionally on a person by person basis doesn’t cause much harm because everything stays the same. To illustrate this: Jack has $10 and Jill has $100. If the money supply were to increase 5% ($110 x .05 = $5.50) and dispersed proportionately to wealth, Jack would now have $10.50 and Jill $105.00. Their relative wealth, ie their claims on other goods and assets, remains unchanged.

The Federal Reserves system does not operate this way. Instead, money is expanded into the hands of select entities (the government and banks) at controlled interest rates. In the above scenario with Jack and Jill, the prices of everything positively correlate with the money supply, including the price of money across time, ie the interest rate. In the new scenario, things are different: Jack has $10 and Jill has $100, but a 5% increase is only given to Jill (at a controlled interest rate). So now, Jack still has $10 while Jill has $105.50. Her access to capital/goods is now 5% greater relative to Jim, an effective transfer of wealth of $.50 from Jack to Jill. Things are more complicated than that because of interest rates and fractional reserve banking, but the essential point to take home is that the first receiver of new money is on the receiving end of a real wealth transfer. As that cash is used to purchase assets and goods by Jill, the prices rise before the cash finally reaches Jack. In other words, the friends of the central bank win at the expense of everyone else.

That this is all facilitated by the manipulation of interest rates by the Federal Reserve leads to false price signals on the market which stimulate malinvestment (or, much less likely, underinvestment). So, even if a taxing system were devised to redistribute all the transferred funds back to Jack, Jack and Jill will both have engaged in activities which orient the structure of production toward less productive activities causing waste.

In sum, the real effect of inflation is a wealth destroying wealth transfer.

Note: Expectations also play a key role. When price increases are expected to occur, they can adjust accordingly, but when Fed policy is secret or poorly understood, it is less likely that such such expectations will be accurate.

In the “nominal” realm, there are problems with price “deflation.” It’s often explained that prices, especially wages, are sticky. What the person who brings this up is trying to convey is that people usually aren’t very well attuned to real wage increases. For example, if prices in general go down 10%, while wages decrease by 5%, people will begin to “under consume” -- behaving as if there had been only a decline in wages when in fact real wages increased. This can be seen as a type of market irrationality.

But this wouldn’t be a one sided problem since price inflation would see similar effects. When wages increase nominally, but not in real terms, as a result of monetary expansion, it’s possible that people would “over consume” -- behaving as if had been an increase in wages when real wages had actually declined. “Deflation” is no excuse for "inflation." (In fact, I view it more as a general attempt to keep demand constantly pushed forward benefiting corporations by disincentivizing saving.)

Is there a way of generating a solution combining all of this information? I think so. Say a yearly basket of goods measures average market price declines of 3% per year. The money supply could be increased by 3% to mirror price declines and be distributed in proportion to each individual’s cash holdings. This way you can have real wage increases without nominal prices causing irrational human purchasing patterns. In other words, real price deflation with zero or low nominal price fluctuations. I’m not sure if this is all really necessary, but it’s worth considering. And all this can be done without a central bank which is completely undesirable in the first place since its entire (dis)function is economically untenable central planning.