From Christy Romer and David Romer:
From the early 1950s to the early 1990s, increases in Social Security benefits in the United States varied widely in size and timing, and were generally not undertaken in response to short-run macroeconomic developments. This paper uses these benefit increases to investigate the macroeconomic effects of changes in transfer payments. It finds a large, immediate, and statistically significant response of consumption to permanent changes in transfers. The effects of temporary benefit changes, in contrast, appear small. The consumption effects of the permanent changes appear to decline at longer horizons, and there is no clear evidence of effects on production or employment. Finally, there is strong evidence of a sharply contractionary monetary policy response to permanent benefit increases, which may account for the apparent decline of the consumption effects and their failure to spread to broader indicators of economic activity.
[...] The most important finding is that transfers matter. Our estimates suggest that a permanent increase in Social Security benefits raises consumer spending in the first month the larger checks arrive almost one-for-one—that is, virtually all of the higher transfers move quickly into higher expenditures. The estimated impact remains near one for about six months, and then gradually declines. The initial impact is highly statistically significant, but the standard errors increase as the horizon lengthens.
We find a marked difference in the response of consumer spending to temporary and permanent benefit changes. While the estimated response to permanent changes is roughly one-for-one and statistically significant, the estimated response to temporary changes is small and not statistically significant. And, for both types of changes, we find no evidence of a response before the payments are actually received. We also fail to find clear evidence of an impact of even the permanent increases on broader macroeconomic outcomes, such as production and employment. [...]
These findings have implications for both research and policy. On the research side, the fact that consumer spending responds strongly when some households receive transfers suggests a failure of Ricardian equivalence. Permanent increases in Social Security benefits raise aggregate consumption even though the tax increases to pay for the higher benefits are often enacted in the very same legislation.
The results are somewhat more supportive of a narrower interpretation of the permanent income hypothesis. The fact that aggregate consumption responds much more to permanent benefit changes than to temporary ones is consistent with the notion that households smooth consumption in response to temporary changes. Because the temporary changes in our sample are typically quite large—and larger than the permanent changes—our findings are consistent with previous work suggesting that the permanent income hypothesis describes household behavior more accurately when larger payments are involved.
At the same time, the fact that consumers appear to respond only to actual increases in benefits, and not in anticipation of them, is less consistent with the permanent income hypothesis. It is possible that consumers were not aware of the impending increases, or that they were liquidity constrained. But their failure to respond to temporary increases argues against the liquidity constraint explanation. So a mystery remains.
Another mystery deserving of further research is why the near-term response to permanent benefit changes is so large. The estimates imply that virtually the entire benefit increase is spent immediately. Perhaps the marginal propensity to consume of the elderly is quite large because they are in the dissaving phase of their life cycle. Alternatively, liquidity constraints or behavioral factors could be important. Determining which of these explanations is correct is important for understanding whether our results for Social Security increases are likely to carry over to other types of transfers.
On the policy side, the fact that we find little impact from large temporary increases in transfers could raise questions about the efficacy of such payments for countercyclical purposes. And, if the findings for temporary transfers carry over to temporary tax changes, our results could raise similar concerns about the countercyclical effectiveness of such tax changes. However, because the temporary transfers that drive our estimates are quite large, our estimates do not speak directly to the issue of whether small temporary transfers could have a greater impact.
Another policy implication of our study involves the interaction of monetary and fiscal policy. We find both statistical and narrative evidence that monetary policymakers in our sample period raised interest rates quickly and substantially in response to permanent increases in Social Security benefits and the related rise in consumer spending. This strong counteracting response presumably muted the expansionary impact of the benefit increases, and may explain our failure to find clear evidence that they affected broader macroeconomic outcomes. These findings confirm that the effects of fiscal policy are very dependent on the conduct of monetary policy. And they suggest that if fiscal policymakers want to achieve some objective, coordination with monetary policymakers may be essential.