1) The document presents a DSGE model that examines how R&D and technology adoption can contribute to business cycle persistence. The model includes endogenous productivity determined by R&D creation of new technologies and the adoption of those technologies.
2) The model is estimated using Bayesian methods and finds that much of the productivity decline following the Great Recession was an endogenous response, particularly through a drop in technology adoption due to the financial crisis and recession.
3) The results suggest that insufficient demand during the Great Recession contributed to the productivity slowdown, and that the mechanism in the model can help account for smooth inflation during that period. The findings indicate that recent low productivity growth may reflect cyclical rather than secular factors.
It discusses growth accounting and Total Factor Productivity (TFP).
This approach is very useful for LDC country in order to have sustained high rate of growth.
It discusses growth accounting and Total Factor Productivity (TFP).
This approach is very useful for LDC country in order to have sustained high rate of growth.
EU ETS innovation fund and low-carbon breakthrough technologiesTomas Wyns
Presentation on the design of forthcoming EU ETS innovation fund. The barriers and options for industrial low-carbon technologies and examples of low-carbon breakthrough technologies in steel, aluminium, chemicals and cement
HLEG thematic workshop on measuring economic, social and environmental resili...StatsCommunications
HLEG thematic workshop on Measuring economic, social and environmental resilience, 25-26 November 2015, Rome, Italy, More information at: http://oe.cd/StrategicForum2015
IM2012 International Conference on Innovation Methods for Innovation Management and Policy - FOREIGN DIRECT INVESTMENT AND PRODUCTIVITY SPILLOVERS: Firm Level Evidence from Chilean industrial sector. Leopoldo Laborda and Daniel Sotelsek.
EU ETS innovation fund and low-carbon breakthrough technologiesTomas Wyns
Presentation on the design of forthcoming EU ETS innovation fund. The barriers and options for industrial low-carbon technologies and examples of low-carbon breakthrough technologies in steel, aluminium, chemicals and cement
HLEG thematic workshop on measuring economic, social and environmental resili...StatsCommunications
HLEG thematic workshop on Measuring economic, social and environmental resilience, 25-26 November 2015, Rome, Italy, More information at: http://oe.cd/StrategicForum2015
IM2012 International Conference on Innovation Methods for Innovation Management and Policy - FOREIGN DIRECT INVESTMENT AND PRODUCTIVITY SPILLOVERS: Firm Level Evidence from Chilean industrial sector. Leopoldo Laborda and Daniel Sotelsek.
1. R&D and Technology Adoption
as
Sources of Business Cycle Persistence
by
Diego Anzoategui, Diego Comin , Mark Gertler and Joseba Martinez
NYU and Dartmouth
November 2015
0
2. Motivation
Slow recovery following a …nancial crisis
– Deleveraging, ZLB and …scal constraints only part of the explanation
Supply-side factors potentially relevant:
– Productivity growth slowdowns follow …nancial crises
True for recent crisis for both U.S. and Euro area data
True following emerging market …nancial crises (Queralto)
Candidate hypotheses: Bad luck versus endogenous response
Bad luck view (at least for U.S): Fernald 2014
– Observes that slowdown in TFP began in 2004, prior to Great Recesion
1
3. Endogenous response to business cycle conditions
Crisis induced large drop in investment in new technologies (both R&D and adop-
tion)
– Large R&D contraction during Great Recession
– Large contraction also in 2001-2002 recession ! TFP decline prior to GR
Speed of technology di¤usion is pro-cyclical.
– Survey data: sample of 26 production technologies that di¤used at various
times over the period 1947-2003 in the US (5) and the UK (21).
– Elasticity of speed of di¤usion with respect to business cycle around 4.
2
6. Figure 3:
patents which suggests that they engage in more R&D activity. They interpret the increase
in the productivity gap as evidence that followers have slowed down the rate at which they
incorporate frontier technology, developed by the leaders.
These co-movement patterns between the business cycle and measures of investments in
technology as well as their outcomes provides, in our view, sufficient evidence to motivate
the quatitative exploration we conduct with the help of our model.
3 Model
Our starting point is a New Keynesian DSGE model similar to Christiano et al. (2005)
and Smets and Wouters (2007). We include the standard features useful for capturing the
data, including: habit formation in consumption, flow investment adjustment costs, variable
capital utilization and ”Calvo” price and wage rigidities. In addition, monetary policy obeys
a Taylor rule with a binding zero lower bound constraint.
The key non-standard feature is that total factor productivity depends two endogenous
variables: the creation of new technologies via R&D and the speed of adoption of these new
technologies. Skilled labor is used as an input for the R&D and adoption processes.
proximately 20%.
9
7. This Paper
Develop and estimate monetary DSGE model with endogenous technology via
R&D and adoption
Use model to assess:
– How much of the recent productivity growth decline was an endogenous re-
sponse to the Great Recession.
– Whether the mechanism can also account for the pre-GR productivity decline.
– More generally, the extent to which endogenous productivity can help account
for business cycle persistence
3
8. Model Features
Core framework: monetary DSGE (Smets/Wouters, CEE)
Non-standard features:
– Endogenous productivity via R&D and technology adoption
– Skilled labor is input for R&D and adoption processes
Endogenous growth due to expanding variety of intermediate goods (a la Romer)
– Aggregate output (to …rst order):
Yt = [A# 1
t t] (UtKt) (Lt)1
– At stock of adopted intermediate goods (technologies)
– ! A# 1
t endogenous part of TFP; t exogenous part
4
9. R&D and Adoption
Endogenous TFP, via At; results from R&D and adoption.
Creation of intermediate goods Zt from R&D.
Zt+1
Zt
= tLsrt +
0 < 1; 0 < < 1:
Free entry condition
! Etf t;t+1Jt+1 tZtL
1
srt g wst = 0
– Jt+1 value of unadopted technology
– Jt+1 procylical and wst sticky ! Lsrt procyclical
5
10. R&D and Adoption (con’t)
Adoption: conversion of Zt to At:
At+1 = t [Zt At] + At
– t = (ZtLsat) probability technology is adopted with 0 > 0; 00 < 0
– ! 1
t
= mean di¤usion lag
Value of adopted good
Vt = mt + Etf t;t+1Vt+1g
FONC for adoption
Zt
0 Etf t;t+1[Vt+1 Jt+1]g = wst
! Vt Jt procylical and wst sticky ! Lsat procyclical
6
11. Liquidity Shocks
Preference shock %t for riskless bonds (zero in net supply)
! foncs for capital and bonds
1 = Etf t;t+1Rkt+1g
1 = Etf t;t+1Rt+1g + t
t = %t=u0(Ct)
! Etf t;t+1(Rkt+1 Rt+1)g = t
Rise in %t reduces both consumption and investment demand
– Also reduces R&D and adoption since t;t+1 1=Rkt+1 declines
– Transmits through economy like monetary shock (shift in Rt+1)
– Increases spread Rkt+1 Rt+1 like …nancial shock
7
12. Estimation Strategy
Conventional Bayesian estimation
Observable variables: Seven standard quarterly series plus R&D
– Standard series: Yt; Ct; It; Wt=Pt; Nt; rn
t ; t
– R&D: spending by private …rms on R&D, including software development
Annual series ! mixed frequency estimation
One shock for each series. Mostly standard except:
– Liquidity premium replaces discount factor as main demand shock
– Shock to productivity of R&D investment
Sample period: 1984Q1-2012Q4
– Structural parameters estimated over non-ZLB period 1984Q1-2008Q4
– Historical decompositions over entire sample (imposing ZLB constraint)
8
18. Concluding Remarks
Estimate DSGE model with endogenous productivity via R&D and adoption
– Use model to identify source of productivity decline following Great Recession
Key result: Much of the productivity decline an endogenous response to recession
– Drop in adoption due to …nancial crisis/recesssion main channel
– Overall, insu¢ cient demand during GR contributed to productivity slowdown
Mechanism also helps account for smoothness in in‡ation during GR
Overall, results suggest that recent low productivity growth may re‡ect (medium
term) cyclical factors as opposed to secular ones.
9