Thom Hartman has conducted an excellent interview with Randall Wray. In part, the interview considers why the currency value (of the US dollar) has declined since the breakdown of Bretton Woods. This led to an interesting post by Mike Norman, who argues that, against a broad basket of currencies, the US dollar has not declined* in value but in fact risen. Both Wray and Norman make valid points, but my understanding is that they are discussing different things. Norman is considering the external value of the currency (i.e. exchange rates), whereas Hartmann and Wray appear to be discussing the domestic value of the currency. It would be possible for the external value of the dollar to have risen against a broad basket of other currencies even though the domestic value of the dollar has declined. However, this latter tendency does not prevent all the dollars in aggregate from purchasing more real output now than in the past. Nor does it prevent a given amount of labor time from commanding more real output than in previous periods.
In the MMT perspective, the domestic value of the currency is the amount of labor time required to obtain a dollar. Specifically, it is taken to be the amount of minimum-wage labor time (or 'simple' labor time) that determines this value. This value goes down over time as nominal wages increase.
As already noted, this does not prevent all the dollars, in aggregate, from buying more real output than in earlier decades. Rising productivity means that a given amount of simple labor time produces progressively more real output in successive periods.
In principle, you could have an economy -- if under competitive conditions -- in which the level of nominal wages, w, was constant and prices, P, fell as productivity increased. Over time, the domestic value of the currency would be constant, because the same amount of labor time would be required to obtain a dollar, but real wages (w/P) and real living standards would be increasing.
This scenario would be debilitating in reality, because deflation causes the real value of debt obligations to rise, a situation that plays even more into the hands of rentiers than the current institutional setup.
In practice, of course, we have rising nominal wages, prices and productivity, which is a good thing. This combination is capable of delivering higher real living standards and higher real wages provided productivity improves. Both these things are possible even though the domestic value of a single dollar will be declining.
There is a connection between the domestic value of the currency and inflation. Inflation will occur if nominal wages and the price markup over wage costs, k, grow faster than the average productivity of labor, APL. The relationships involved are explained in an informative post by Eric Tymoigne. For present purposes, it is enough to note that:
P* = k* + w* - APL*
The stars indicate that the terms refer to growth rates of the variables involved. (Usually dots above the variables would be used.) The expression implies that a decline in the value of the currency (a rise in nominal wages) will be associated with inflation unless productivity improves rapidly enough to offset both this decline in the value of the currency and any widening of the markup.
This suggests that one reason -- probably the major reason -- that we have inflation is due to tussling over distribution. Workers may correctly mistrust the degree of competition in product markets, and so seek nominal-wage increases rather than waiting for a fall in prices. At the same time, particular segments of workers may be trying to increase their wages relative to other workers and people on fixed incomes. More importantly in an environment of weak organized labor, capitalists with market power seek to increase the markup whenever they can get away with it. In any of these cases, inflation is the way that incompatible claims on real income are resolved.
Unfortunately for workers, the weakening of organized labor over the neoliberal period has suppressed real-wage growth. To the extent incomes of working families have increased in real terms, it is largely because of the increase in two-income households.
* Matias Vernengo has since provided an interesting analysis indicating that since 1971 the U.S. dollar exchange rate, though appreciating against a broad basket of currencies in nominal terms (Mike Norman's observation), has depreciated against that same basket in real terms. This does not alter the basic point of the present post, which is concerned with the domestic value of the currency.