Routine-Biased Technological Change and HoursWorked over the Business Cycle

Get Complete Project Material File(s) Now! »

The transatlantic employment gap

In this section, I account for cross-country diﬀerences in employment outcomes over time. First, I compare the average socio-demographic composition and employment propensities in France and the U.S. Then, I grasp the extent to which those components account for the employment gap over time. Furthermore, I identify the occupational nature and the socio-demographic structure of the French employment deficit. I further identify the extent to which the French employment deficit translates into unemployment and non-participation.

Socio-demographic composition and propensities

Diﬀerences in employment outcomes arise from discrepancies either in the socio-demographic com-position of the working age population or in employment propensities. Thus, I document the aver-age propensities and socio-demographic composition to roughly grasp similarities and discrepancies across countries. The demographic composition of a country is captured by the fraction of type g individuals in the total working age population g;t. The propensities of type g individuals to be in the state Sg;t is defined as the fraction of the type g working age population in the considered state.
Table 1.1 documents the average demographic composition as well as propensities to be employed by task for each aggregate socio-demographic group in both countries.10 On average, there are significant cross-country diﬀerences in the educational composition, as well as in propensities of young and old individuals to be employed. For instance, France is only composed by 19.78% of individuals with at least a high school degree against 30.45% for the U.S. The propensity to be employed for the most qualified group is approximately 20 pp higher than for the less qualified group in both countries. Furthermore, despite a similar age structure, the propensity to be employed across diﬀerent age groups varies significantly. It is much lower for young and old individuals in France compared to the U.S. with 30.72% and 34.93% in France and 50.83% and 57.17% in the U.S.
On the contrary, employment propensities of prime-aged individuals are slightly higher in France (79.11%) than in the U.S. (78.23%). Moreover, both countries have a similar gender and age compo-sition but diﬀerent educational structures. The working age population of both countries is composed of approximately 51% of women and 49% of men. It contains roughly 63% of prime-aged workers against around 21% of young and 16% of old individuals. At first sight, transatlantic employment diﬀerences are emanating from discrepancies in employment propensities by age and in the skill composition of the working age population.

The transatlantic employment gap

I now describe the transatlantic employment gap. I quantify the importance of socio-demographic and propensity discrepancies in accounting for the employment gap. I proceed by decomposing the employment gap across time, occupational and socio-demographic groups. In doing so, I reassess the long-run structural analysis initiated by Piketty (1998) and determine the occupational nature of the French employment deficit.
Employment per capita in country c at time t (Etc) is a by-product of propensities and sizes of socio-demographic groups. Hence, it can be written as G Xg Etc = g;tcEg;tc (1.9).

Decomposition by socio-demographic groups

I now describe how these composition and propensity discrepancies are distributed across socio-demographic groups. I show that they are concentrated in some key groups.
Figure 1.3 displays the contribution of each socio-demographic group to the employment gap for discrepancies in both composition (left panels) and employment propensities (right panels) over time. In 1982, the employment deficit induced by employment propensity discrepancies (Gpro) is negligible. It is mainly accounted for by a lack in manual employment from young low-skilled individuals. They account for 3.41 pp of the propensity component (0.83 pp), from which 62.80% arise from manual occupations against 27.61% and 9.59% from routine and abstract occupations, respectively. The lower employment outcomes of French young low-skilled individuals are partly oﬀset by the higher employment outcomes of prime-aged workers. They account for -4.16 pp of the propensity component, from which 81.47% arise from an excess in French abstract employment propensities.
In 1998, all socio-demographic groups contribute to the deterioration of French employment outcomes relatively to the U.S. Nevertheless, it originates primarily from a deficit in routine jobs for low-skilled young and senior individuals. Their contribution adds up to a total of 8.04 pp. This represents 72.78% of the employment deficit from which 63.70% stem from a deficit in routine jobs against 22.34% and 13.66% from manual and abstract jobs, respectively. Thus, out of the 4.1 million job deficit in 1998, 3.0 million concern low-skilled young and senior individuals out of which 1.9 million stem from a lack in routine occupations. In 2017, the employment deficit is still significant but falls sharply. French low-skilled young and senior individuals still suﬀer from lower employment prospects with respect to the U.S. They contribute for 4.37 pp to the employment deficit from which 56.84% arise from routine occupations against 24.29% and 18.87% from manual and abstract occupations, respectively.

Aggregate relevance of socio-demographic groups

Despite displaying diﬀerent employment dynamics across periods, both countries experience abstract and manual employment gains and routine employment losses at some point in time. I assess the aggregate importance of socio-demographic groups in accounting for occupational employment dynamics.
Figure 1.5 depicts employment change decompositions for each socio-demographic group for the overall 1982 to 2017 period.14 Abstract employment increases by 7.75 pp in France between 1982 and 2017. Prime-aged individuals with more than a high school degree account for most of this increase through their rising group size which reflects a rise in skill supply. They contribute for 8.60
pp. In the U.S., the overall rise in abstract employment is of 9.46 pp. Prime-aged individuals with more than a high school degree contribute for 6.87 pp. Nevertheless, prime-aged women with more than a high school degree contribute twice as much as prime-aged men with similar education levels. It is also noteworthy that senior individuals with more than a high school degree also contribute to the abstract employment rise in both countries but to a much lesser extent.
Routine employment falls dramatically in both country between 1982 and 2017. It decreases by -9.88 pp in France and by -8.32 pp in the U.S. Routine employment losses are mostly endured by low-skilled young and prime-aged individuals. For those groups, the fall in routine employment occurs both through deteriorating employment prospects and the drop in their group size. In France, those groups contribute for a -16.96 pp decline in routine employment. Roughly -9.67 pp arise from the declining propensity to work in such jobs and roughly -10.42 pp from the declining group size. Comovements in propensities and group sizes dampen the fall by approximately 3.12 pp. In the U.S., they contribute for -11.46 pp in the decline in routine employment from which -5.84 pp arise from the propensity component and -7.22 pp from the composition shifts. Comovements in propensities and group sizes dampen this fall by 1.60 pp. Such patterns reflect the consequences of task-biased technological change, globalization as well as the shift in the skill composition of the workforce. In both countries, prime-aged high-skilled individuals slightly dampen the fall in routine employment with a contribution of 5.36 pp and 2.21 pp in France and the U.S., respectively. This is in line with Beaudry, Green, and Sand (2016) who argue that there has been a reversal in the demand for skills in the 2000s in the U.S. that pushes high-skilled workers down the occupational ladder towards jobs traditionally performed by lower-skilled workers.
Manual employment rises in both countries between 1982 and 2017. It increases by 3.66 pp in France and 2.71 pp in the U.S. The rise in manual employment is primarily accounted for by low-skilled prime-aged individuals in both countries with a contribution of 2.00 pp in France and 1.04 pp in the U.S. The rise in manual employment propensities is especially strong for women reflecting their rising labor market participation over the period. Prime-aged high-skilled workers also contribute to the increase in manual employment but to a much lesser extent in France (0.88 pp) compared to the U.S (1.60 pp). This contribution of prime-aged high-skilled individuals highlights their occupational downgrading.
Thus, prime-aged individuals are at the core of the polarization of occupational employment over the last four decades. However, low-skilled workers suﬀer severe employment losses and only benefit from limited employment gains. Thus, labor market policies should focus on promoting low-skilled employment by expanding labor demand as France did since the mid 1990s with its labor cost re-duction policies targeted on low-paid workers. However, the sustainability of the employment gains arising from those policies remains partly threatened by further technological and trade develop-ments. In that respect, the development of training programs might facilitate the transition of low-skilled workers towards occupations that oﬀer more stable employment prospects. Furthermore, policies should increase incentives for low-skilled young and senior workers to participate in the labor market as they account for most of the transatlantic employment deficit.

READ computational higher-order logics

Occupational wage dynamics

According to Autor and Dorn (2013), job polarization captures the rise in manual and abstract labor demand and the decline in routine labor demand. Such reallocation process led to a polarization of workers compensation in the U.S. Those data are not available for France since only net wages are observable in the FLFS. They do not polarize and even reveal a decline in occupational net wage inequalities. I argue that this is not due to socio-demographic compositional shifts but rather to changing occupational net wage schedules.

The environment

There are three main sectors which are perfectly competitive: a goods sector, a service sector and a non-market sector. They use three inputs which are unskilled labor, skilled labor and capital. Those inputs are used to accomplish manual, routine and abstract tasks. Therefore, each worker is characterized by a set of skills fa; r; mg with a, r and m referring to abstract, routine and manual tasks, respectively. There is a unit mass of skilled labor which is only used to perform abstract tasks
64 Chapter 2. Job Polarization and Unskilled Employment Losses in France and thus characterized by the set of skill levels f1; 0; 0g.6 There is a unit mass of unskilled labor that can accomplish two types of tasks, i.e. routine and manual tasks. Unskilled workers have the same ability to accomplish manual tasks. However, they are heterogeneous with respect to their ability to perform routine tasks, which is captured by the eﬃciency parameter 2 [0; +1[ with density function f( ) = e as in Autor and Dorn (2013). Consequently, each unskilled worker is characterized by a skill set f0; ; 1g. Capital can also perform routine tasks.

Table of contents :

1 Transatlantic Employment Outcomes
1.1 Introduction
1.2 Data
1.2.1 Labor force surveys and samples
1.2.2 Variables and measurement
1.2.3 Break correction
1.3 Aggregate employment and job polarization
1.3.1 Aggregate employment outcomes
1.3.2 Job polarization
1.4 The transatlantic employment gap
1.4.1 Socio-demographic composition and propensities
1.4.2 The transatlantic employment gap
1.4.3 Decomposition by socio-demographic groups
1.4.4 Non-employment
1.5 Employment dynamics
1.5.1 Employment change decomposition
1.5.2 Aggregate relevance of socio-demographic groups
1.6 Labor market transitions
1.6.1 Average transition rates
1.6.2 Job polarization and occupational mobility
1.7 Conclusion
2 Job Polarization and Unskilled Employment Losses in France
2.1 Introduction
2.2 Stylized facts
2.2.1 The deterioration of unskilled employment outcomes
2.2.2 Job polarization
2.2.3 Occupational wage dynamics
2.2.4 Labor taxation policies
2.2.5 An incomplete reallocation of unskilled labor
2.3 A general equilibrium model
2.3.1 The environment
2.3.2 Production
2.3.3 Occupational choice
2.3.4 The representative consumer
2.3.5 Market clearing conditions
2.3.6 Equilibrium
2.4 Calibration
2.5 Results
2.5.1 The obstructed reallocation of unskilled labor
2.5.2 Labor taxation and technological change
2.5.3 Accounting for the decline in unskilled employment
2.5.4 Discussion
2.6 Conclusion
3 Routine-Biased Technological Change and HoursWorked over the Business Cycle
3.1 Introduction
3.2 A general equilibrium model
3.2.1 The model
3.2.2 Calibration
3.2.3 Comparative statics
3.3 Data
3.3.1 Data construction
3.3.2 Stylized facts
3.4 A VAR model
3.4.1 Bayesian estimation
3.4.2 Identification
3.5 Results
3.5.1 Specification I – Is Gali’s technological shock neutral?
3.5.2 Specification II – RBTC, neutral and task-supply shocks
3.5.3 Technological shocks and aggregate fluctuations
3.6 Conclusion
Conclusion
A Appendix of chapter 1
A.1 Alternative measure of the employment deficit
A.2 Margin-error adjustment
A.3 Additional Tables and Figures
B Appendix of chapter 2
B.1 Additional Figures and Tables
B.2 Data
B.2.1 French Labor Force Survey and samples
B.2.2 Building variables
B.3 Re-weigthing methods
B.3.1 Counterfactual employment structure
B.3.2 Wage change decompositions
B.4 Labor taxation policies
B.4.1 Labor taxation time series
B.4.2 A brief history of labor taxation policies
B.5 Asymptotic equilibrium
B.5.1 Preliminary computations
B.5.2 Asymptotic wages
B.5.3 Asymptotic allocation of labor
B.5.4 Asymptotic wage inequality
C Appendix of chapter 3
C.1 Comparative statics analysis
C.2 Univariate time series analysis
C.2.1 Unit root tests
C.2.2 Robustness of business cycle moments
C.3 VAR algorithm
C.4 Additional results
C.4.1 Impulse responses
C.4.2 Forecast error variance decompositions
C.4.3 Historical decompositions
C.5 Empirical robustness
List of Figures
List of Tables
Résumé français
Asbtract