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