Intertemporal empirics Flashcards
RW?
Hall 1978.
Assume u’ linear (remove prudence) and so C(r,Yp)
Excess sensitivity?
Flavin, 1981.
Reaction to known Y changes not 0.
Liquidity constraint
Campbell and Mankiw 1989 s the model
1980s US =0.2 (Hall and Mishkin, 1982) / Gao and Kim 2014 =0.85 in China.
Possibility of future LC also matters (Hayashi 1987).
However, Shea 1995 suggests corr persists even in households with lots of liquid assets?!
Expenditure on C =/= C?
Aguiar and Hurst, 2005 consider food diaries in retirement and find time spent shopping / preparing up!
Excess smoothness?
C should immediately react to Yp thus should be just as variable. Campbell and Deaton (1989) ratio of variance =0.64! Further, Yp is in fact more variable than current Y they argue!
Near rational in smoothness?
Large (>10%) Y change consistent with PIH (Paxson, 1993). ‘near rational’ Parker, 1999.
Consider costs of changing consumption (Browning and Collada, 2001).
Habits (Constantinides, 1991)
Prudence?
Non linear u’. Can show that this means ‘excess smoothness’ in face of uncertainty desirable (Attanasia, 1999)
Especially relevant in young people
Ricardian equivalence central idea?
A neo-classical consumer will save all of a transitory tax cut (Tnpv)
Support for RE?
Carroll and Summers, 1987: Budget deficits move 1:1 with private savings (US/Canada).
Cheney: Reagan proved deficits did not matter
Against RE: LC
Rahman et al (2013): Tax cuts stimulated private consumption in Malaysia
Tax effects on distribution (violates Ricardian Equivalence)
Present taxpayers =/= beneficiaries? (Bernheim, 1989)
Tax distortionary?
Derive EE with a proportional tax rate. Smooth taxation desirable (Barro, 1979): Less distortionary. Prescott, 2004: This is primary reason for differing Ls across countries (US higher LS)
RBC: Volatile investment?
Abel et al 1988: Change in I is roughly 90% of variation of AD
RBC: Procyclical Ls to prod/wage.
Is this prediction supported by evidence?
No. Ls moves more to w, not w1/w2 as it should based upon EE! (Altongi, 1982).
Underpredicts Var(employment) (Stadler,1994): Inelastic Ls and market imperfections.
Mankiw, 1989: Requires ppl willing to sub leisure over time?!
RBC: Capital Utilization rate?
Using the Solow residual to estimate prod assumes constant K utilization rate! But, procyclical (Bernanke and Rotenburg, 1995) eg labour hoarding (Fay and Medoff, 1985). Indicates lower prod
Is the RBC prediction of Procyclical prod supported?
Fits sign, overpredicts magnitude: only equal to 30-60% delta Y (Burnside et al 1993)
Importance of AD?! (Summers, 1986)
Also, be wary of using Solow residual (see capital utilization rate)
RBC: Persistence of output fluctuations
Tech innovations persist. Cambell and Mankiw: Shocks ARE followed up by larger movements. Although this doesn’t say anything on why / if it reverts LR!
RBC consider RW
See all the RW evidence
Conclusion to RBC
Doesn’t hold well enough to be considered to hold! Thus, policy makers should smooth Y and or other variables
Micro-founded: Kenynes was wrong
Micro-founded uses the classical story (Lucas and Sargent 1979)
UK: IE vs SE
IE dominates (Blundell et al 2010
Life cycle approach / PIH
Modigliani 1986 / Friedman
Estimates of depreciation rate
Where is this useful?
4-15% for machines / 2-4% for buildings (Blanchard, 2013).
Volatile investment in RBC! This is because replacing depreciating capital is not so large relative to new capital requirements
NBC: r pro/ counter
Perma: Pro - US 1990s to early 2000s
Temp: Eg Denslow and Rush 1989: Fall of French r due to harvest shortfalls
Liquidity constraint evidence for?
Coibion et al 2020: LC consumers spent 10% more of transfer by CARES act
New Ricardian Equivalence
Barro, 1974
TNPV matters
Government debt represents an asset and liability (Ricardo, 1820)
Evidence of wars and fiscal policy
Barro 1997: Wars in US (eg Vietnam, Korea, WW) show Y up, C down, LS up in war (consistent with temporary FP prediction)
Evidence that income effect dominates in LR
US weekly manufacturing hours 60 to 42 1890-1996 (Doppelhofer, 2009)
LCH evidence against
Prudence and bequests to children mean we don’t observe LCH (Browning and Crossley, 2001)
Poor likely to go on benefits so don’t have incentive to smooth C?
Use benefits as insurance! Moral Hazard issue (Hubbard et al, 1994). Large proportion of population with virtually no savings: heterogeneity of savings motives (asset based means testing).
Thus, we see individuals consuming entirety of current income
Accumulation of wealth due to bequests important so PIH>LCH
Kotlikoff and Summers (1981)
Closer to a ‘representative agent for family’ than individual, self-interested individuals
RBC prediction: Pro-cyclical wages
Only mildly procyclical, corr(w, Y) = 0.12 (King and Rebelo, 1999)
Wealthy hand to mouth?
Cost of accessing illiquid wealth too high so just consume income (Parker et al, 2013)
Hysteresis DF test?
Strongest form of hysteresis (Ue as random walk) rejected (Lee and Chang, 2008)