Accelerator and Multiplier effects at full employment

Question

I have a question regarding the multiplier and accelerator effects. The way I see it, the value of the multiplier tends to be lower as the economy approaches full capacity. I would therefore expect growth to be slower as full capacity is reached. The accelerator effect works best when there is no excess capacity as an increase in aggregate demand would mean producers use their already existing idle fixed capital.

However the accelerator is based on past increases in consumption and output; for the investment to increase, growth must therefore increase at an an increasing rate. If at long run full capacity, growth is slower due to a lower multiplier value and the inability of firms to expand output, how can the accelerator effect come into existence???

I am sorry that this was long but I am unsure as to whether I have made a mistake early in my "chain of reasoning"...

Answer 1

Grigory, I am glad you are skeptical of this 'theory'.

I am not sure I understand you line of argument correctly, but given the assumptions it seems to be correct. Accelerator cannot be sustained in long-run due to fixed factors of production (think land, labor force, etc). Accelerator theory is the sense you are presenting it is not present in modern macroeconomic literature or graduate curriculum.

Let's then turn our attention to the assumption that 'multiplier' at full capacity. The old Keynesian theory indeed assumed that the multiplier is higher when there is high unemployment. Indeed, this was a reason for advocating for public spending during recessions.

However, modern evidence do not seem to support this view. For example Ramey & Zubairy (2018, Journal of Political Economy), show that the multiplier does not vary significantly with the business cycle if one controls for expectations. Keep in mind, however, that those things are notoriously hard to measure. An ideal experiment, that would allow us, economists to measure the multiplier, would consist of an unexpected debt-financed government spending in the absence of central bank. You can see that all of these assumptions are typically violated (taxes and central bank's action matter in particular).

So the bottom line is, always take simple, undergraduate-level models with a pinch of salt. Some are logically inconsistent and most (if not all) are consistent with data.