This is a development of a comment I made here. It is my attempt to summarise, using my language rather than his, what Scott Sumner has to say on economic stimulus and why we need to focus on nominal GDP (i.e. GDP in ordinary dollar terms). It includes accepting his re-statement of part of my original comment in the second-last paragraph.
NGDP = nominal GDP = total monetary value of productive (as in ‘counts to GDP’) transactions. An excellent post on the patterns of NGDP in the US and the OECD since 1960 is here. (The astonishing fall in nominal GDP since 2008, its biggest fall since the 1930s, can be seen quite clearly.)
RGDP = real GDP = total production of goods and services (i.e. NGDP adjusted by some price deflator).
Money is how one makes and accepts offers. (“I am willing to spend x on this”, “I am willing to sell this for y”.) It is therefore the fundamental economic signalling device.
The more smoothly supply can respond, the more any rising level of money offers generates more output. The less smoothly supply can respond, the more any rising level of money offers generates higher prices. This is true in any particular market but also in the aggregation across all markets.
Inflation is the degree to which money offers increases systematically and persistently beyond the output response across all markets.
But if folk think their money will speak more loudly later than now (i.e. the value of money is expected to rise as prices are expected to fall), they will hold on to it more, leading to a lowering of aggregate money offers and so of the level of transactions/economic activity. (Hence the contractionary effects of deflation.)
If folk think their money will speak less loudly later (i.e. the value of money is expected to fall as prices are expected to rise), they will tend to spend it more, leading to an increase in aggregate money offers. (Hence the stimulatory effects of inflation.)
When considering (monetary or fiscal) stimulus, RGDP is misleading because stimulus doesn’t directly affect RGDP. It affects NGDP (aggregate money spent) with RGDP (aggregate goods and services produced) responding according to conditions on the supply side.
To put it another way, stimulus operates by raising money offers above what they otherwise would have been* but the effects of any such increase in money offers on total output depends on the supply response. Hence the concern with how far under existing productive capacity an economy is running and with the efficiency of its supply responses in responding to demand either within existing productive capacity or by expanding productive capacity.
* Even if the stimulus money is just saved, it can still have a stimulatory effect if increased confidence means that people spend at a higher level than they otherwise would have. I.e. money that would have been saved is spent instead. Of course, if the stimulus fails to improve people's expectations about their income prospects, then there will be little effect.
Musings On Iraq In The News - My interview with Charlie Winter on how the Islamic State is dealing with its defeat in Mosul was reprinted by Vox Pol.
2 minutes ago