Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
EuropeEurope’’s Productivity Growth s Productivity Growth Slumps But Employment Surges: Slumps But Employment Surges:
Why? Why? Ian DewIan Dew--Becker, NBERBecker, NBER
Robert J. Gordon, Northwestern and NBERRobert J. Gordon, Northwestern and NBERPresented at Presented at CESifoCESifo Economic Economic Studies ConferenceStudies Conference onon
Productivity and GrowthProductivity and GrowthMunich, 22 June 2007Munich, 22 June 2007
Ian in SF, you canIan in SF, you can’’t see t see ““MV=PYMV=PY””
This is a Work in ProgressThis is a Work in Progress
At the end IAt the end I’’ll tell you some of our plans for further ll tell you some of our plans for further researchresearchTodayToday’’s Presentation Combines a Joint Version from s Presentation Combines a Joint Version from Last September with a Solo Version of IanLast September with a Solo Version of Ian’’s from s from FebruaryFebruaryOne Thing we have in Common: Loud ColorsOne Thing we have in Common: Loud ColorsSince his first day as my RA 3.5 years ago, he has come Since his first day as my RA 3.5 years ago, he has come up with inspired color schemes, like everything up with inspired color schemes, like everything involving EU must be yellowinvolving EU must be yellow--blue and US must be redblue and US must be red--whitewhite--blueblueOccasional lapses here toward black and whiteOccasional lapses here toward black and white
The US Accelerates,The US Accelerates,Europe DeceleratesEurope Decelerates
From 1950 to 1995 EU productivity growth was faster From 1950 to 1995 EU productivity growth was faster than in the USthan in the USBut in the past decade since 1995 we have witnessedBut in the past decade since 1995 we have witnessed
An explosion in US productivity growthAn explosion in US productivity growthA slowdown in EU productivity growth roughly equal in sizeA slowdown in EU productivity growth roughly equal in sizeAn explosion in research on the US takeoff and but much less An explosion in research on the US takeoff and but much less research on Europeresearch on Europe’’s slowdowns slowdown
The magnitude of the shift (average EKS&GK Groningen)The magnitude of the shift (average EKS&GK Groningen)EU/US level of labor productivity (ALP)EU/US level of labor productivity (ALP)1979 1979 1995 1995 2004200480%80% 97%97% 89%89%
Point of Departure: PostPoint of Departure: Post--95 95 Turnaround Plus New HeterogeneityTurnaround Plus New Heterogeneity
This paper begins with two simple observations:This paper begins with two simple observations:1. While European productivity (Y/H) has fallen back 1. While European productivity (Y/H) has fallen back
since 1995 relative to the US, output per capita since 1995 relative to the US, output per capita (Y/N) has not fared nearly as badly(Y/N) has not fared nearly as badly
►►Y/H growth gap: .9%Y/H growth gap: .9%►►Y/N growth gap: .2%Y/N growth gap: .2%
2. After 1995, we see divergence across the EU2. After 1995, we see divergence across the EU--15 in 15 in Y/H growthY/H growth
►► St. Dev. 1970St. Dev. 1970--1995: 0.621995: 0.62►► St. Dev. 1995St. Dev. 1995--2005: 1.012005: 1.01
The Key Identity SuggestsThe Key Identity Suggeststhe Tradeoffthe Tradeoff
An identity links Y/N and Y/H to H/N:An identity links Y/N and Y/H to H/N:Y/N = Y/H * H/NY/N = Y/H * H/N
Thus the paradox of high European Y/H and low Y/N must be Thus the paradox of high European Y/H and low Y/N must be resolved by lower H/Nresolved by lower H/NAlso, Y/H and H/N are jointly determinedAlso, Y/H and H/N are jointly determinedThe task of this paper is going to be figure out which The task of this paper is going to be figure out which direction the causation runsdirection the causation runs
We will argue that a good deal of the decline in ALP growth We will argue that a good deal of the decline in ALP growth is due to exogenous employment shocksis due to exogenous employment shocksAlso we will highlight the reversal of almost everything at Also we will highlight the reversal of almost everything at 1995, comparing 19701995, comparing 1970--95 vs. 199595 vs. 1995--20052005
Bringing Together the Disparate Bringing Together the Disparate LiteraturesLiteratures
Literature #1, why did EuropeLiterature #1, why did Europe’’s hours per capita s hours per capita (hereafter H/N) decline before 1995? Prescott, (hereafter H/N) decline before 1995? Prescott, RogersonRogerson, , SargentSargent--LundqvistLundqvist, , AlesinaAlesina, Blanchard, Blanchard
High taxes, regulations, unions, high minimum wagesHigh taxes, regulations, unions, high minimum wagesEurope made labor expensiveEurope made labor expensiveMovement up Labor Demand curve => low employment + Movement up Labor Demand curve => low employment + high ALPhigh ALP
Literature #1 has missed the turnaroundLiterature #1 has missed the turnaroundSince 1995 there has been a decline in tax rates and Since 1995 there has been a decline in tax rates and employment protection measures; unionization earlieremployment protection measures; unionization earlierBig increase in hours per capita, turnaround in both absolute Big increase in hours per capita, turnaround in both absolute terms and relative to the US Move back down Lterms and relative to the US Move back down LDD curvecurve
Textbook Labor EconomicsTextbook Labor Economics
-2
-1
0
1
2
3
4
5
6
7
1 2 3 4 5 6 7 8 9 10 11
Labor Input
Rea
l Wag
e
Labor Demand Curve
High-Cost LaborSupply Curve
Low-Cost LaborSupply Curve
(W/P)0
(W/P)1
N0 N1
Downward shift in labor supply curve reduces real wage and productivity
A
B
PrePre--1995: Moving Northwest1995: Moving Northwest
19701970--95 EU climbs to the northwest95 EU climbs to the northwestHours per capita decline, average labor productivity Hours per capita decline, average labor productivity increasesincreasesIn this sense much of EuropeIn this sense much of Europe’’s 1970s 1970--95 productivity 95 productivity catchupcatchup was was ““artificial,artificial,”” propelled by policies making propelled by policies making labor expensivelabor expensive
No busboys, grocery baggers, valet parkersNo busboys, grocery baggers, valet parkersProduct regulations kept stores shut tight many hours of Product regulations kept stores shut tight many hours of the day/nightthe day/nightAll this reduced EuropeAll this reduced Europe’’s employment share in s employment share in retail/services retail/services
PostPost--1995: Moving Southeast1995: Moving Southeast
19951995--2004 EU slides southeast2004 EU slides southeastHours per capita start increasing while they decline in the USHours per capita start increasing while they decline in the USEffects are magnified by slow reaction of capital, eventually Effects are magnified by slow reaction of capital, eventually capital should grow faster offsetting much or all of capital should grow faster offsetting much or all of productivity slowdownproductivity slowdown
Literature #1 misses the turnaroundLiterature #1 misses the turnaroundSince 1995 decline in tax rates and employment protection Since 1995 decline in tax rates and employment protection measuresmeasuresWe are unaware of much macroWe are unaware of much macro--level research on the level research on the turnaround in hoursturnaround in hoursAllard and Allard and LindertLindert (2006) do not really mention it (2006) do not really mention it –– data only data only goes to 2001goes to 2001
Literature #2: The EULiterature #2: The EU--US ALP gapUS ALP gap
Central Focus of Lit #2 on postCentral Focus of Lit #2 on post--1995 1995 turnaround in US Productivity Growthturnaround in US Productivity Growth
Jorgenson, Ho and Jorgenson, Ho and StirohStiroh (2006): (2006): ’’9595--’’00 due to ICT, 00 due to ICT, ’’0000--’’05 something else05 something elseRetail is often notedRetail is often noted
Van Ark, Van Ark, InklaarInklaar and and McGuckinMcGuckin (2003)(2003)Foster, Foster, HaltiwangerHaltiwanger and and KrizanKrizan (2002) on new (2002) on new establishmentsestablishments
BailyBaily and and KirkegaardKirkegaard (2004) on regulations(2004) on regulationsNeed to free land use restrictionsNeed to free land use restrictions
Fully 85% of EU productivity slowdown has its Fully 85% of EU productivity slowdown has its counterpart in a speedcounterpart in a speed--up of EU H/Nup of EU H/N
Europe paid for lower ALP mainly with higher Europe paid for lower ALP mainly with higher hours rather than less consumptionhours rather than less consumptionSaltariSaltari and and TravagliniTravaglini have made a similar point with have made a similar point with respect to Italyrespect to Italy
This runs counter to the Blanchard story about This runs counter to the Blanchard story about preferences for leisurepreferences for leisure
Now we hear that theyNow we hear that they’’re not lazy, just unproductivere not lazy, just unproductiveHuge literature on different structural reasons for Huge literature on different structural reasons for EU sclerosisEU sclerosis
Literature #3: relationship between Literature #3: relationship between Y/H and H/NY/H and H/N
There is a long line of research examining the There is a long line of research examining the relationship between hours and productivityrelationship between hours and productivityEven using an IV approach, increases in H/N drive Even using an IV approach, increases in H/N drive down Y/Hdown Y/H
This makes sense in a single factor model or with any slow This makes sense in a single factor model or with any slow adjustment of capitaladjustment of capitalMeasuring the speed of adjustment of investment is difficult Measuring the speed of adjustment of investment is difficult –– future research for usfuture research for us
View todayView today’’s talk as a report on research in progress, s talk as a report on research in progress, not the final polished wordnot the final polished word
Figure 1. Trends in Output per Hour, Output, and Hours, U.S. and EU, Anual Growth Rates, 1970‐2005
0
1
2
3
4
5
6
1970 1975 1980 1985 1990 1995 2000 2005
Perc
ent
E.U. Output per Hour
U.S. Output per Hour
Interpreting the PostInterpreting the Post--1995 1995 TurnaroundTurnaround
Simple HP trendsSimple HP trendsEurope is continuing its long slow declineEurope is continuing its long slow declineTurnaround is generally pegged at 1995Turnaround is generally pegged at 1995
The EUThe EU--15 stops catching up, and the US takes off15 stops catching up, and the US takes offWe are mainly going to examine the We are mainly going to examine the determinants of the turnaround determinants of the turnaround –– i.e. changes in i.e. changes in Y/H growth postY/H growth post--19951995Qualification: US trend peaks in 2002Qualification: US trend peaks in 2002--03 and is 03 and is now decliningnow declining
New US Productivity Trends BasedNew US Productivity Trends Basedon March 2007 Quarterly Dataon March 2007 Quarterly Data
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
NFPB LP
Total economy LP
Difference
-2
-1
0
1
2
3
4
1970 1975 1980 1985 1990 1995 2000 2005
Perc
ent
US Output per CapitaEU‐15 Output per Capita
EU‐15 Hours per Capita
US Hours per Capita
We Need to Look at EverythingWe Need to Look at EverythingPer CapitaPer Capita
Population growth in EU 0.7 percent per year Population growth in EU 0.7 percent per year slower than US over the past decadeslower than US over the past decadeOutput per capita in the EU doesnOutput per capita in the EU doesn’’t look bad at t look bad at allallPostPost--1995 hours turnaround is a counterpart to 1995 hours turnaround is a counterpart to the Y/H turnaroundthe Y/H turnaroundWe will see that there is a similar pattern We will see that there is a similar pattern withinwithinthe EU the EU –– strong negative correlation between strong negative correlation between the hours and ALP turnaroundsthe hours and ALP turnarounds
-2
-1
0
1
2
3
4
5
6
7
1984 1988 1992 1996 2000 2004
US Hours
US CapitalEU Capital
EU Hours
The US has experienced an enormous decline in The US has experienced an enormous decline in hours growth when capital growth fellhours growth when capital growth fell
Thus Thus ““capitalcapital--deepeningdeepening”” numbers for US are numbers for US are misleading as they reflect as much movements in the misleading as they reflect as much movements in the denominator as in the numerator.denominator as in the numerator.Cumulative hours growth zero 2000Cumulative hours growth zero 2000--06, growth in 06, growth in hours per capita negativehours per capita negative
The EU had strong hours growth while the US The EU had strong hours growth while the US went through its recession and recoverywent through its recession and recovery
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
1984 1988 1992 1996 2000 2004
US TFP
EU TFP
Defining Tigers and Tortoises, Defining Tigers and Tortoises, Pop Shares and Private ALP GrowthPop Shares and Private ALP Growth
Tigers: Ireland, Finland, GreeceTigers: Ireland, Finland, GreecePop Share: 5%Pop Share: 5% ALP 4.79%ALP 4.79%
Middle: Sweden, Austria, UK, Germany, Middle: Sweden, Austria, UK, Germany, Portugal, FrancePortugal, France
Pop Share: 61%Pop Share: 61% ALP: 2.45%ALP: 2.45%Tortoises: Belgium, Netherlands, Denmark, Tortoises: Belgium, Netherlands, Denmark, Luxembourg, Spain, ItalyLuxembourg, Spain, Italy
Pop Share: 34%Pop Share: 34% ALP: 0.72% ALP: 0.72%
1970‐1995 1995‐2005 Difference 1970‐1995 1995‐2005 Difference 1970‐1995 1995‐2005 Difference
US 1.42 2.30 0.88 0.55 -0.14 -0.69 1.97 2.15 0.18EU 2.89 1.40 -1.49 -0.80 0.55 1.35 2.09 1.95 -0.14Tigers 2.93 2.95 0.02 -0.67 1.22 1.89 2.26 4.17 1.91Middle 2.80 1.86 -0.94 -0.84 -0.08 0.76 1.96 1.78 -0.19Tortoises 3.05 0.39 -2.66 -0.75 1.59 2.34 2.30 1.98 -0.32
Growth RatesGrowth RatesGrowth RatesProductivity Hours per Capita Output per Capita
We break the EU-15 into three groups based on post-’95 Y/H growth:
Tigers: Ireland, Finland and Greece Middle Countries: Sweden, Austria, UK, Germany, Portugal and France Tortoises: BeNeLux, Denmark, Spain and Italy
A closer look at the TortoisesA closer look at the Tortoises
Mainly driven by Spain and ItalyMainly driven by Spain and ItalySpain:Spain:
►►--4.44% turnaround in Y/H4.44% turnaround in Y/H►►+5.01%+5.01% turnaround in H/Nturnaround in H/N
Italy:Italy:
►►--2.25%2.25% turnaround in Y/Hturnaround in Y/H►►+1.08% turnaround in H/N+1.08% turnaround in H/N
Had we ranked the countries according to Had we ranked the countries according to output per capita, Spain would be a Tiger output per capita, Spain would be a Tiger
Figure 2. Private Economy Labor Productivity Growth by Country: 1979‐1995, 1995‐2003
1995-20031979-1995
0 1 2 3 4 5 6 7 8 9
Italy
Spain
Luxembourg
Denmark
Netherlands
Belgium
France
Portugal
Germany
United Kingdom
Austria
Sweden
Greece
Finland
IrelandTigers
Middle
TortoisesPre-1995
Post-1995
Making Sense of CrossMaking Sense of Cross--EUEUHeterogeneityHeterogeneity
Notice the homogeneity preNotice the homogeneity pre--1995 and heterogeneity 1995 and heterogeneity postpost--’’9595The only two countries with a noticeable acceleration The only two countries with a noticeable acceleration are Greece and Irelandare Greece and IrelandSweden a bit up and UK a bit downSweden a bit up and UK a bit downSharp declines for France, Portugal, and all the Sharp declines for France, Portugal, and all the TortoisesTortoisesFor most of the remainder of the paper, we focus only For most of the remainder of the paper, we focus only on the middle countries and tortoiseson the middle countries and tortoises
The tigers are special cases The tigers are special cases –– they do not provide any policy they do not provide any policy lessons for the rest of the EUlessons for the rest of the EU
The New Results in thisThe New Results in thisPaper at the Industry LevelPaper at the Industry Level
We aggregate productivity growth by industry in a way We aggregate productivity growth by industry in a way that allows us to determine the relative role of that allows us to determine the relative role of productivity and sharesproductivity and sharesThe The ““productivityproductivity”” effect is just the difference in effect is just the difference in productivity growth in a given industryproductivity growth in a given industryThe The ““shareshare”” effect is the addition or subtraction from effect is the addition or subtraction from growth as shares shift within industries.growth as shares shift within industries.
Example: Ireland shifts to high tech manufacturing, this Example: Ireland shifts to high tech manufacturing, this comes out as a comes out as a ““shareshare”” effect within manufacturingeffect within manufacturing
Contributions, Productivity vs. Share Contributions, Productivity vs. Share Effects, in EUEffects, in EU--US, 1995US, 1995--20032003
-0 .7 -0 .6 -0 .5 -0 .4 -0 .3 -0 .2 -0 .1 0 0 .1 0 .2
Farm s/m in ing
C onst./utilities
M anufacturing
R eta il/w holesale
T rans.
F inance
S erv .
C om m .
R ea l es tate
P rodS hare
N on-IC T share
N on-durables share
N on-IC T prodIC T prod
N on-durables prod
IC T share
Manufacturing is nearly as importantas retail
But ICT is tinyOnly ~2% hours share
ALP growth multiplied by nominal sharesALP growth multiplied by nominal shares
-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Real Estate
Communications
Services
Finance
Transportation
Retail/Wholesale
Manufacturing
Construction Utilities
Farms/Mining
U.S.
E.U.
US acceleration is widespread, not just in retailand manufacturing.
EU weakness is also widespread
Tortoises vs. MiddleTortoises vs. Middle
-0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1
Farms/mining
Const./utilities
Manufacturing
Retail/wholesale
Trans.
Finance
Serv.
Comm.
Real estate
Share
Prod
Failure is more widespread.Totally unrelated industries account for the declineNote that this is largely driven by productivity, not share effects
Interpreting the TortoiseInterpreting the TortoiseProblem after 1995Problem after 1995
Failure is across the boardFailure is across the boardConsistent with basic theme of paper, that there is a Consistent with basic theme of paper, that there is a macro causemacro cause
How much due to a reduction in taxes and in regulations?How much due to a reduction in taxes and in regulations?How much remains for an exogenous decline in TFP growth?How much remains for an exogenous decline in TFP growth?
Understanding Share EffectsUnderstanding Share EffectsICT Share higher in US ICT Share higher in US vsvs EU and also middle EU and also middle vsvs tortoisestortoisesBig EU share deficit in retail/wholesale and services, Big EU share deficit in retail/wholesale and services, consistent with high tax storyconsistent with high tax story
Part of Tiger success is moving resources, out of Part of Tiger success is moving resources, out of agriculture for Greece and Ireland, into ICT mfg for agriculture for Greece and Ireland, into ICT mfg for Ireland and FinlandIreland and Finland
Research StrategyResearch Strategy
Divergence across the EU has increasedDivergence across the EU has increasedThe Y/H slowdown in the tortoises in most The Y/H slowdown in the tortoises in most countries is balanced by healthy H/N growthcountries is balanced by healthy H/N growthWe are going to then try to break down the We are going to then try to break down the determinants of the middledeterminants of the middle--tortoise gap in Y/H tortoise gap in Y/H growth and relate it to H/N growthgrowth and relate it to H/N growth
Figure 4. Employment per Capita
90
95
100
105
110
115
120
125
1983 1988 1993 1998 2003
Tortoises
Middle Countries
EU
Hours per Employee
84
86
88
90
92
94
96
98
100
102
1983 1988 1993 1998 2003
Tortoises
Middle Countries
EU
Interpreting the Graphs ofInterpreting the Graphs ofE/N and H/EE/N and H/E
(H/N) = (E/N) * (H/E)(H/N) = (E/N) * (H/E)’’7979--’’95 US minus EU H/N growth: 1.01%95 US minus EU H/N growth: 1.01%
Half from employment per capita (E/N), half from hours per Half from employment per capita (E/N), half from hours per employee (H/E) employee (H/E) US had rising E/N, EU had falling H/EUS had rising E/N, EU had falling H/E
’’9595--’’04, gap was 04, gap was --.76% (EU had higher growth).76% (EU had higher growth)E/N gap was E/N gap was --.85%, H/E .09%.85%, H/E .09%Almost entirely explained by a shift up in EU E/NAlmost entirely explained by a shift up in EU E/NH/E seems to have stabilizedH/E seems to have stabilized
So when comparing employment to ALP, E/N is the So when comparing employment to ALP, E/N is the margin we are going to focus onmargin we are going to focus on
0
0.2
0.4
0.6
0.8
1
1.2
15 to 19 20 to 24 25 to 34 35 to 44 45 to 54 55 to 64 65+
EU-15Tortoises
Middle Countries
E/N Ratio to the US
-- A lot is explained around 45-54 and 15-19-- All are very similar for 35-44
Figure 7. Difference in Growth Rates of Employment per Capita by Sex‐Age Group, Tortoises minus Middle Countries, 1995‐2005 minus 1985‐1995, Employment and Share Effects
Employment Share
-1 -0.5 0 0.5 1 1.5 2
Women 65+
Women 55‐64
Women 45‐54
Women 35‐44
Women 25‐34
Women 15‐24
Men 65+
Men 55‐64
Men 45‐54
Men 35‐44
Men 25‐34
Men 15‐24
Contributions to the difference in the turnaround in the Middle countries versus the Tortoises
This is the standard shiftThis is the standard shift--share analysis from share analysis from industryindustry--level productivity studies (see level productivity studies (see StirohStirohand van Ark and and van Ark and InklaarInklaar))Note that the Tortoises have a big passive Note that the Tortoises have a big passive advantage advantage –– share effects for 25share effects for 25--3434Large employment effects for prime age womenLarge employment effects for prime age women
Slightly smaller for prime age menSlightly smaller for prime age men
Teens and retirement aged contribute littleTeens and retirement aged contribute little
Male and Female employment rates
Notice the enormous growth in female E/NIt even manages to have the biggest acceleration following 1995
Men in the Tortoises have caught up, women still have a long way to go
Average Growth RatesMiddle 1985 1995 2005 85‐95 95‐05 turnaroundMale 65.85 62.30 60.79 ‐0.55 ‐0.25 0.31Female 41.46 44.81 48.09 0.78 0.71 ‐0.07
Tortoises 1985 1995 2005 85‐95 95‐05 turnaroundMale 57.72 57.93 60.94 0.04 0.51 0.47Female 26.02 30.97 39.88 1.74 2.53 0.79
Variables to explain E/NVariables to explain E/N
Tax wedgeTax wedgeEPL EPL –– measures of bargaining coordination, firing measures of bargaining coordination, firing restrictions, etc.restrictions, etc.Percentage of employees part timePercentage of employees part time
Actually see little evidence of the business cycleActually see little evidence of the business cycleWe can see whether part time employees are new entrants to We can see whether part time employees are new entrants to the labor forcethe labor force
Union densityUnion densityUnion density and union power arenUnion density and union power aren’’t the samet the sameFrance has always had lower union density than the US France has always had lower union density than the US
Explanatory variables are the tax wedge, EPL, Explanatory variables are the tax wedge, EPL, union density and net reservation wageunion density and net reservation wage
Net reservation wage measures generosity of Net reservation wage measures generosity of unemployment benefitsunemployment benefits
We donWe don’’t worry about factors affecting teens or t worry about factors affecting teens or those near retirement because those age groups those near retirement because those age groups dondon’’t drive much of the divergence within the t drive much of the divergence within the EUEU
Figure 5. Tax Wedge
20
25
30
35
40
45
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
Tortoises
EU
Middle Countires
Recall PrescottRecall Prescott’’s claim that the entire gap s claim that the entire gap between EU and US employment can be between EU and US employment can be explained by tax wedgesexplained by tax wedgesIf tax wedges are the main drivers of If tax wedges are the main drivers of employment variation, the compression in EU employment variation, the compression in EU taxes is interestingtaxes is interesting
►►Policy and E/N are converging but Y/H is Policy and E/N are converging but Y/H is divergingdiverging
Employment Protection Legislation (EPL)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
EU
Middle Countries
Tortoises
Figure 6. Union Density
20
25
30
35
40
45
1960 1965 1970 1975 1980 1985 1990 1995 2000
Middle Countries
Tortoises
EU
Net Reservation Wage
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
1960 1965 1970 1975 1980 1985 1990 1995 2000
Middle Countries
Tortoises
EU
Interpreting the GraphsInterpreting the Graphsof the Explanatory Variablesof the Explanatory Variables
EPL shows the same convergenceEPL shows the same convergenceUnion density shows the familiar declineUnion density shows the familiar decline
This is a messy variable because union power is This is a messy variable because union power is criticalcriticalThe US has more unions than FranceThe US has more unions than France
The net reservation wage has risen, with the The net reservation wage has risen, with the Tortoises converging up rather than downTortoises converging up rather than down
Regressions of employment per capitaPopulation weighted, US and Lux. excludedNotice the importance of fixed effectsNet reservation wage and EPL have positive coefficients
VariableTax Wedge -0.51 *** 0.01 -0.68 *** -0.30 ***EPL -0.01 0.10 ***Union Density -0.23 *** 0.15 ***Output Gap 1.12 * 1.88 ** 0.79 1.42 *Net Reservation Wage 0.10 *** 0.06 ***
R2 0.59 0.01 0.66 0.23RMSE 0.135 0.205 0.122 0.181Number of Observations 352 352 352 352Fixed Effects? yes no yes no
E/N regressions by age, FE includedNote the effect of the output gap declines with age (see Jaimovich)Tax wedge has smaller effect on men and prime age workersUnion density almost always has negative effects
Gender Age R2Men 15‐24 -1.02 *** 0 -0.05 *** 2.95 *** 0.1 ** 0.81Women 15‐24 -1.03 *** 0.02 -0.04 * 2.5 *** 0.14 *** 0.88Men 25‐34 -0.23 *** 0.01 -0.02 *** 1.26 *** -0.02 * 0.66Women 25‐34 -0.43 *** 0.13 *** 0.08 *** 1.14 ** -0.07 * 0.74Men 35‐44 -0.26 *** 0 0.01 0.73 *** -0.04 *** 0.53Women 35‐44 -0.8 *** 0.13 *** 0.19 *** 0.56 -0.28 *** 0.82Men 45‐54 -0.5 *** -0.03 0.09 *** 0.25 -0.21 *** 0.49Women 45‐54 -0.93 *** 0.08 0.23 *** 0.2 -0.54 *** 0.8Men 55‐64 -0.43 *** -0.07 ** -0.11 *** 0.77 * 0.19 *** 0.82Women 55‐64 -0.67 *** 0 0.01 0.81 * -0.15 *** 0.95Men 65+ -1.26 *** 0.08 -0.47 *** -2.1 0.15 0.78Women 65+ -1.34 *** 0.07 -0.42 *** -1.83 0.18 0.75
Gap DensityTax Wedge EPL WageNet. Res. Output Union
Next We Turn to the PossibleNext We Turn to the PossibleTradeoff of Y/H vs. E/NTradeoff of Y/H vs. E/N
We next run regressions of productivity growth We next run regressions of productivity growth on employmenton employment
See Gordon(1997), See Gordon(1997), BeaudryBeaudry and Collard (2001), and Collard (2001), McGuckinMcGuckin and van Ark (2005), basically any 1and van Ark (2005), basically any 1--factor factor modelmodel
Even with instruments, the relationship is robust Even with instruments, the relationship is robust across countries and time periodsacross countries and time periods
BeaudryBeaudry and Collard provide evidence that the and Collard provide evidence that the coefficient has shifted over timecoefficient has shifted over time
Regressions of Productivity on Employment
Instruments are explanatory variables from prior regressions
Variable LagsEmployment Rate 0 -0.59 *** -0.52 ***
1 -0.09 -0.15 -0.05 -0.0210 -0.07 -0.05
Sum of all Lags -0.81 *** -0.68 *** Standard Error [0.13] [0.13]Sum of Lags 1 and 2 -0.69 *** -0.62 *** Standard Error [0.1] [0.09]
Change in Output Gap 0.82 *** 0.78 ***Ratio to US LP -0.022 *** -0.041 ***
Fixed Effects? no yes
Comments on the ProductivityComments on the ProductivityRegressionsRegressions
Coefficient on employment is Coefficient on employment is --.7 to .7 to --.8.8No bounce back with later lagsNo bounce back with later lagsSignificant catchSignificant catch--up effectup effect
Being 10% behind the US adds .2Being 10% behind the US adds .2--.4% to ALP .4% to ALP growth each yeargrowth each year
Country fixed effects do not affect results much, Country fixed effects do not affect results much, as opposed to employment regressionsas opposed to employment regressions
We can now ask how policy shifts affected We can now ask how policy shifts affected productivity growthproductivity growth
This is very much back of the envelope This is very much back of the envelope –– we need to we need to be more careful in the futurebe more careful in the future
Two basic effectsTwo basic effectsPolicy effectPolicy effectFemale cultural effectFemale cultural effect
We canWe can’’t identify the total cultural effect on t identify the total cultural effect on women; we just get the gap the middle countries women; we just get the gap the middle countries and tortoises:and tortoises:
Take residual male employment growthTake residual male employment growthCall MiddleCall Middle--Tortoise gap the endogenous partTortoise gap the endogenous part
To get exogenous female growth, take the MiddleTo get exogenous female growth, take the Middle--Tortoise gap for female residuals, and subtract the Tortoise gap for female residuals, and subtract the endogenous effectendogenous effectBasically, female residual growth minus male residual Basically, female residual growth minus male residual growth equals cultural effectsgrowth equals cultural effectsWe can consider alternative identifying assumptions: We can consider alternative identifying assumptions: get the B functions from regressionsget the B functions from regressions
Excess employment growth in the Tortoises
Using the above methodology, we get excess female growth of .63% per yearExcess policy driven employment growth of .13%Note the massive overprediction for US employment growth
Short digression on US trends and forecasts
Actual Predicted Actual Predicted Actual Predicted ResidualUS 62.89 61.23 62.34 70.74 -0.10 1.61 -1.70Middle 53.22 53.37 53.96 55.16 0.15 0.37 -0.21Tortoises 44.02 43.33 49.58 45.31 1.32 0.50 0.82
1995 2004 Avg. Growth Rate
Breaking Down the MiddleBreaking Down the Middle--Tortoise GapTortoise Gap
►►.13% gap in predicted .13% gap in predicted ΔΔE/NE/N→→.1% gap in Y/H.1% gap in Y/H
►►.63% excess female E/N growth.63% excess female E/N growth→→.48% gap in Y/H.48% gap in Y/H
Adding the two exogenous employment shocks Adding the two exogenous employment shocks and multiplying by .75 gives a predicted shortfall and multiplying by .75 gives a predicted shortfall of .58%of .58%Of the 1.47 percentage point gap, we can Of the 1.47 percentage point gap, we can explain 38% with employment effectsexplain 38% with employment effects
Should we expect this to continue?Should we expect this to continue?Women in the Tortoises still need to raise Women in the Tortoises still need to raise employment by 8% to catch up to the middle employment by 8% to catch up to the middle countriescountries
Translates to a 7.7% total gapTranslates to a 7.7% total gapImplies a further 5.75% shortfallImplies a further 5.75% shortfallOver ten years would imply a shortfall of .58% per Over ten years would imply a shortfall of .58% per yearyear
Increased investment would offset some of thisIncreased investment would offset some of this
ConclusionsConclusions
Across Europe we find a negative correlation Across Europe we find a negative correlation between employment and productivity growthbetween employment and productivity growthAs labor markets have been liberalized, some As labor markets have been liberalized, some countries have experienced huge rises in countries have experienced huge rises in employmentemploymentExogenous shocks can explain about 40% of the Exogenous shocks can explain about 40% of the shortfall in ALP in the tortoisesshortfall in ALP in the tortoisesFuture research needs to identify the sources of Future research needs to identify the sources of the other 60%, starting with the other 60%, starting with
a return to the industrya return to the industry--byby--industry analysisindustry analysisA dynamic analysis of capital adjustmentA dynamic analysis of capital adjustment