- (2016): âThe great reversal in the demand for skill and cognitive tasks,â Journal of Labor Economics, 34(S1), S199âS247.
Paper not yet in RePEc: Add citation now
- ⢠Cohort 2004. Beginning: 1 = November 2003; End: 98 = December 2011.
Paper not yet in RePEc: Add citation now
Acemoglu, D. (1999): âChanges in unemployment and wage inequality: An alternative theory and some evidence,â American economic review, 89(5), 1259â1278.
- Acemoglu, D., and D. Autor (2011): âSkills, tasks and technologies: Implications for employment and earnings,â in Handbook of labor economics, vol. 4, pp. 1043â1171. Elsevier.
Paper not yet in RePEc: Add citation now
- Allman, E. S., C. Matias, and J. A. Rhodes (2009): âIdentifiability of Parameters in Latent Structure Models with Many Observed Variables,â Annals of Statistics, 37(6A), 3099â30132.
Paper not yet in RePEc: Add citation now
Arcidiacono, P., and J. B. Jones (2003): âFinite Mixture Distributions, Sequential Likelihood and the EM Algorithm,â Econometrica, 71(3), 933â946.
- Arcidiacono, P., E. Aucejo, A. Maurel, and T. Ransom (2016): âCollege Attrition and the Dynamics of Information Revelation,â Working Paper 22325, National Bureau of Economic Research.
Paper not yet in RePEc: Add citation now
Arcidiacono, P., V. J. Hotz, A. Maurel, and T. Romano (2020): âEx Ante Returns and Occupational Choice,â Journal of Political Economy, 128(12), 4475â4522.
- Argan, D., and R. J. Gary-Bobo (2021): âLes diplômes français se sont-ils dévalorisés?,â Revue Economique, forthcoming.
Paper not yet in RePEc: Add citation now
Ashworth, J., V. J. Hotz, A. Maurel, and T. Ransom (2021): âChanges across cohorts in wage returns to schooling and early work experiences,â Journal of labor economics, 39(4), 931â964.
- At this point the beginning and the end of each spell plus the observations at the time of the survey are kept as observations of the individual. Each row of the database becomes an observation (i, t) in the labor market of an individual i (either employed or not), at a date t. At this point the number of observations are : 171,258 in 1998, 133,211 in 2004, 91,174 in 2010.
Paper not yet in RePEc: Add citation now
- At this point, the dataset can be described as follows: ⢠Cohort 1998: 15,950 individuals that are observed on average 10.74 times; (minimum 1, 1st quartile 6; median 10; 3rd quartile 14; maximum 54) ⢠Cohort 2004: 12,233 individuals that are observed on average 10.89 times; (minimum 1, 1st quartile 6; median 10; 3rd quartile 14; maximum 63) ⢠Cohort 2010: 8,774 individuals that are observed on average 10.39 times; (minimum 1, 1st quartile 6; median 9; 3rd quartile 13; maximum 45) Then, we build the experience variable as the sum of working time up to time t â 1. For each spell we add the information regarding the accumulated experience at time t â 1 at the beginning, and the end of the spell. Now, using the individual dataset we create the variables: father is a professional, place of residence at grade 6 entry and the education level (i.e., degree category). The detail for these variables, for each cohort, can be found in the tables below.
Paper not yet in RePEc: Add citation now
Autor, D. H., L. F. Katz, and M. S. Kearney (2008): âTrends in US wage inequality: Revising the revisionists,â The Review of economics and statistics, 90(2), 300â323.
Autor, D., C. Goldin, and L. F. Katz (2020): âExtending the Race between Education and Technology,â AEA Papers and Proceedings, 110, 347â51.
- Beaudry, P., D. A. Green, and B. M. Sand (2014): âThe declining fortunes of the young since 2000,â American Economic Review, 104(5), 381â86.
Paper not yet in RePEc: Add citation now
- Beffy, M., D. Fougère, and A. Maurel (2012): âChoosing the Field of Study in Postsecondary Education: Do Expected Earnings Matter?,â The Review of Economics and Statistics, 94(1), 334â 347.
Paper not yet in RePEc: Add citation now
Belzil, C., and J. Hansen (2020): âReconciling changes in wage inequality with changes in college selectivity, using a behavioral model,â Institut Polytechnique de Paris.
Blundell, R., D. A. Green, and W. Jin (2022): âThe U.K. as a Technological Follower: Higher Education Expansion and the College Wage Premium,â Review of Economic Studies, 89, 142â180.
- Bonhomme, S., K. Jochmans, and J.-M. Robin (2016): âEstimating multivariate latentstructure models,â Annals of Statistics, 44(2), 540â563.
Paper not yet in RePEc: Add citation now
- Bouveyron, C., G. Celeux, T. B. Murphy, and A. E. Raftery (2019): Model-based clustering and classification for data science: with applications in R, vol. 50. Cambridge University Press.
Paper not yet in RePEc: Add citation now
Card, D., and T. Lemieux (2001): âCan falling supply explain the rising return to college for younger men? A cohort-based analysis,â The quarterly journal of economics, 116(2), 705â746.
Carneiro, P., and S. Lee (2011): âTrends in quality-adjusted skill premia in the United States, 1960-2000,â American Economic Review, 101(6), 2309â49.
- Carneiro, P., J. J. Heckman, and E. J. Vytlacil (2011): âEstimating Marginal Returns to Education,â American Economic Review, 101(6), 2754â81.
Paper not yet in RePEc: Add citation now
Carneiro, P., K. T. Hansen, and J. J. Heckman (2003): â2001 Lawrence R. Klein Lecture Estimating Distributions of Treatment Effects with an Application to the Returns to Schooling and Measurement of the Effects of Uncertainty on College Choice*,â International Economic Review, 44(2), 361â422.
- Cassagneau-Francis, O. (2021): âThe Returns to Higher Education by Cognitive and Noncognitive Abilities,â Discussion paper, UCL, London.
Paper not yet in RePEc: Add citation now
- Cassagneau-Francis, O., R. Gary-Bobo, J. Pernaudet, and J.-M. Robin (2021): âA Nonparametric Finite Mixture Approach to Difference-in-Difference Estimation, with an Application to On-the-job Trainig and Wages,â Discussion paper, Sciences Po, Paris.
Paper not yet in RePEc: Add citation now
- Corblet, P. (2022): âThe Decreasing Returns to Experience for Higher Education Graduates in France: An Occupational Analysis,â Discussion paper, Sciences Po, Paris.
Paper not yet in RePEc: Add citation now
Cunha, F., and J. J. Heckman (2007): âIdentifying and Estimating the Distributions of Ex Post and Ex Ante Returns to Schooling,â Labour Economics, 14(6), 870â893, Education and Risk.
Cunha, F., J. Heckman, and S. Navarro (2005): âSeparating uncertainty from heterogeneity in life cycle earnings,â Oxford Economic Papers, 57(2), 191â261.
Dustmann, C., and C. Meghir (2005): âWages, experience and seniority,â The Review of Economic Studies, 72(1), 77â108.
- Each individual enters the dataset the month after the end of his(her) education. There is a date system for each cohort. Beginning is the date when an individual in the cohort can be first observed, while End is the date of the last observation of the dataset: ⢠Cohort 1998. Beginning: 1 = January 1998; End: 96=December 2005.
Paper not yet in RePEc: Add citation now
- Each of the three stacked surveys contains 3 files: employment spells, non employment spells and individual characteristics, the three files form a dataset containing the sequence of employment and unemployment (or non employment) spells for each individual during 7 years Changes in working hours during employment spells are described. In 1998, the employment-spells dataset contains 47,936 observations, the unemployment dataset contains 30,329 observations and 52 The complete cohort-by-cohort results are available upon request.
Paper not yet in RePEc: Add citation now
Emmons, W. R., A. H. Kent, L. Ricketts, et al. (2019): âIs college still worth it? The new calculus of falling returns,â Federal Reserve Bank of St Louis Review, 101(4), 297â329.
- F Online Appendix: Impact of the Business Cycle. Variations of the National Unemployment Rate Figure 11: Unemployment rate; France, 1992-2019 Cohort 1998 Cohort 2004 Cohort 2010 7 8 9 10 11 Unemployment rate 1 9 9 1 -T 4 1 9 9 2 -T 4 1 9 9 3 -T 4 1 9 9 4 -T 4 1 9 9 5 -T 4 1 9 9 6 -T 4 1 9 9 7 -T 4 1 9 9 8 -T 4 1 9 9 9 -T 4 2 0 0 0 -T 4 2 0 0 1 -T 4 2 0 0 2 -T 4 2 0 0 3 -T 4 2 0 0 4 -T 4 2 0 0 5 -T 4 2 0 0 6 -T 4 2 0 0 7 -T 4 2 0 0 8 -T 4 2 0 0 9 -T 4 2 0 1 0 -T 4 2 0 1 1 -T 4 2 0 1 2 -T 4 2 0 1 3 -T 4 2 0 1 4 -T 4 2 0 1 5 -T 4 2 0 1 6 -T 4 2 0 1 7 -T 4 2 0 1 8 -T 4 2 0 1 9 -T 4
Paper not yet in RePEc: Add citation now
- Figure 10: Men. High-School Degree and Less 1998 2010 2004 20 40 60 80 1200 1300 1400 1500 1600 Comparison of returns to education and experience of male workers with an educational achievement lower than or equal to high-school graduation i.e., the French baccalauréat in three 7-year Generation surveys, 1998, 2004, 2010.
Paper not yet in RePEc: Add citation now
- Figure 9: Men. Masters and âSchoolsâ Degrees 1998 2004 2010 20 40 60 80 1800 2000 2200 2400 2600 2800 Comparison of returns to education and experience of male workers holding Masterâs, business and engineering school degrees in three 7-year Generation surveys, 1998, 2004, 2010. Months of potential experience are on the x-axis; monthly real wages (2013 euros) are on the y-axis. 39 Appendix ?? shows that the picture is somewhat different for young women, whose salaries resisted devaluation much better: after 5 years (i.e., 60 months), the 1998 curve catches up the 2010 curve. In Appendix ??, Fig. ?? shows that the less educated women did not experience any devaluation during our 20 year period.
Paper not yet in RePEc: Add citation now
- Gary-Bobo, R. J., M. Goussé, and J.-M. Robin (2016): âGrade retention and unobserved heterogeneity,â Quantitative Economics, 7(3), 781â820.
Paper not yet in RePEc: Add citation now
- Goldin, C., and L. F. Katz (2008): The race between education and technology. harvard university press.
Paper not yet in RePEc: Add citation now
Guvenen, F. (2007): âLearning Your Earning: Are Labor Income Shocks Really Very Persistent?,â American Economic Review, 97(3), 687â712.
Guvenen, F., F. Karahan, S. Ozkan, and J. Song (2021): âWhat Do Data on Millions of U.S. Workers Reveal About Lifecycle Earnings Dynamics?,â Econometrica, 89(5), 2303â2339.
- Hall, P., and X.-H. Zhou (2003): âNonparametric estimation of component distributions in a multivariate mixture,â The Annals of Statistics, 31(1), 201 â 224.
Paper not yet in RePEc: Add citation now
Heckman, J. J., and B. Singer (1984): âA Method for minimizing the impact of distributional assumptions in econometric models for duration data,â Econometrica, 52, 271â320.
Heckman, J. J., and E. Vytlacil (2005): âStructural equations, treatment effects, and econometric policy evaluation 1,â Econometrica, 73(3), 669â738.
Heckman, J. J., J. E. Humphries, and G. Veramendi (2018): âReturns to Education: The Causal Effects of Education on Earnings, Health, and Smoking,â Journal of Political Economy, 126(S1), S197âS246.
- Ichino, A., A. Rustichini, and G. Zanella (2022): âCollege Education, Intelligence and Disadvantage: Policy Lessons from the UK, 1960-2004,â .
Paper not yet in RePEc: Add citation now
- In contrast, the share of individuals with less than a high-school degree has decreased, but their average log-wage has increased. In general, we observe that the average real monthly wage of fulltime employees increases with education. More interestingly, we remark that there is an increase in the range of the share of individuals working full-time across education degrees. Whereas in 1998, the share of individuals working full time varied from 62% (low education level) to 68% (high education level), in 2010, the same share varied from 46% among the individuals without a high-school degree to 76% for individuals who graduated from a business (or engineering) school.
Paper not yet in RePEc: Add citation now
- Jeong, H., Y. Kim, and I. Manovskii (2015): âThe price of experience,â American Economic Review, 105(2), 784â815.
Paper not yet in RePEc: Add citation now
Kambourov, G., and I. Manovskii (2009): âOccupational specificity of human capital,â International Economic Review, 50(1), 63â115.
Kasahara, H., and K. Shimotsu (2009): âNonparametric Identification of Finite Mixture Models of Dynamic Discrete Choices,â Econometrica, 77, 135â175.
- Katz, L. F., and K. M. Murphy (1992): âChanges in relative wages, 1963â1987: supply and demand factors,â The quarterly journal of economics, 107(1), 35â78.
Paper not yet in RePEc: Add citation now
Magnac, T., and D. Thesmar (2002): âIdentifying Dynamic Discrete Decision Processes,â Econometrica, 70(2), 801â816.
Magnac, T., N. Pistolesi, and S. Roux (2018): âPost-Schooling Human Capital Investments and the Life Cycle of Earnings,â Journal of Political Economy, 126(3), 1219â1249.
- McLachlan, G., and D. Peel (2000): Finite Mixture Models, Wiley series in probability and statistics: Applied probability and statistics. Wiley.
Paper not yet in RePEc: Add citation now
- the individualsâ file contains 16,040 observations. The corresponding figures are 39,101, 22,724, and 12,365 in the 2004 survey; these figures are respectively 26,056, 16,467, and 8,882 in the 2010 survey. In each survey, we start by removing the employment spells that are labelled as family help (i.e., aide familial or afa), self-employed (i.e., Ã son compte or asc), undescribed summer jobs (i.e., vac).
Paper not yet in RePEc: Add citation now
- The real salary is computed in July 2013 euros.
Paper not yet in RePEc: Add citation now
- Then, we merge the employment and non-employment data sets: each individualâs history appears with a sequence of employment and non-employment spells. In 1998 we have 75,117 spells, in 2004 58,253 spells, in 2010 40,467 spells. Individuals are interviewed at the end of their 3rd, 5th and 7th year. They are asked to describe their recent history and their situation at the very moment of the call. So, for each individual, we have 3 additional observations that are the description of their situation at the month of the interview. We recover this information from the 3rd and 5th year of each cohort (i.e., survey) for the individuals observed at the end of the 7th year and we add these data to the 7th year survey.
Paper not yet in RePEc: Add citation now
- Then, we remove individuals lacking an observation of the fatherâs occupation and of the residence at grade 6 entry. This leads us to delete 1,071 individuals in the 1998 cohort, 647 individuals in the 2004 cohort and 856 individuals in the 2010 cohort. Finally, we take the subset of males. The final dataset for each cohort includes 16,404 individuals, among which: ⢠Cohort 1998: 80,006 observations for 7,383 individuals; ⢠Cohort 2004: 60,907 observations for 5,500 individuals; ⢠Cohort 2010: 37,489 observations for 3,521 individuals.
Paper not yet in RePEc: Add citation now
- This amounts to removing 3,148 employment spells in 1998, 3,572 employment spells in 2004, 2,076 employment spells in 2010. It follows that an individual who is always self-employed (or categorized as afa, or vac) in the first 7 years after having left the educational system disappears from the data.
Paper not yet in RePEc: Add citation now
- This increases the number of point observations in each cohort, that, at this point are: 29,986 in 1998, 23,011 in 2004, 16,153 in 2010.
Paper not yet in RePEc: Add citation now
- to 6% per year. The OLS returns to potential experience seem to be only slightly biased (if we compare the estimated coefficients with the corresponding fixed-effects coefficients). In contrast, as expected, the OLS returns to effective experience are biased upwards, and all the more since the attainment level is high. Returns to experience typically increase with the education level, in all cohorts. But the most important feature of Table 17 is that returns to experience fell between 1998 to 2004, and they fell more for the highest degrees of attainment.
Paper not yet in RePEc: Add citation now
- Valletta, R. G. (2018): âRecent flattening in the higher education wage premium: Polarization, skill downgrading, or both?,â in Education, skills, and technical change: Implications for future US GDP growth, pp. 313â342. University of Chicago Press.
Paper not yet in RePEc: Add citation now
- Verdugo, G. (2014): âThe great compression of the French wage structure, 1969â2008,â Labour Economics, 28, 131â144. A Descriptive Statistics Descriptive statistics. Table 9 presents the descriptive statistics of our sample. The share of individuals who graduated from university (with a master degree), from a business or an engineering school has increased substantially. We also observe that their average real log wage has decreased.
Paper not yet in RePEc: Add citation now
- We deleted the employment spells that lack the working time information; as a consequence, we lose 413 observations in 1998, 66 observations in 2004 and 1,536 observations in 2010.
Paper not yet in RePEc: Add citation now
- We stack the three cohorts and generate a unique dataset. We generate a cohort variable c taking values 1998, 2004 or 2010, and a common calendar for the three cohorts where 1 = January 1998 and 240 = December 2017. Table 26 lists the degree types that have been aggregated in each of the categories used for estimation.
Paper not yet in RePEc: Add citation now