Page 44 - IDEA Study 3 2015 Duchodova studie
P. 44

We also simulated the impact of an increasing level of educational attainment in the Czech population in future decades, as built into the current demographic structure. We found that the direct impact of a rise in educational attainment on the employment rate and retirement age profiles is very small; a much greater role is played by the interaction of age and the institutional setup of statutory retirement ages. This is in line with our findings concerning the Czech-Swedish level of educational attainment's low impact of the on the employment rate and the retirement profiles. GRAPH 5.5  Real CZ2007 Simulated CZ2007 Real CZ2013 11 .75 .75 .5 .5 11 .25 .25 .75 .75 00 Real CZ2007 Real CZ2013 Simulated CZ2007                        Real CZ2007 Real CZ2013                    Simulated                      CZ2007                                                                          Real CZ2007   Real CZ2013 Simulated CZ2007                                                           50 55 60 65 70 50 55 60 65 70 .5 .5 Proportion WORKING by age: CZ MEN Proportion WORKING by age: CZ WOMEN GRA .25 1 Real CZ2007 .25 Real CZ2013 Simulated CZ2007 Real CZ2007 Real CZ2013 55 Proportion WORKING by age: CZ WOMEN PH 5.6 1 00 .7550 55 60 65 Proportion WORKING by age: CZ MEN .5 Real CZ2007 Simulated CZ2007 Real CZ2013 70 .7550 .5 60 65 70 Real CZ2007 Real CZ2013 Simulated CZ2007 Simulated CZ2007         11 .25 .25 .75 .75   05       56                         06                        57 Proportion R                         ETIRED by a         ge: CZ WOM N                                              00 050    50 5 .5 .5 .25 .25 E   Proportion 56 RETIRED by 06 57   age: CZ MEN                                          00 50 55 60 65 70 50 55 60 65 70 Proportion RETIRED by age: CZ MEN Proportion RETIRED by age: CZ WOMEN Notes on Data and Regressions Our set of explanatory variables contains about 50 variables, most of them indicator variables 1/0. Since we are not interested in identifying causal effects (this is methodologically impossible) nor in the partial impact of particular variables, we do not deal with possible multicollinearity among our explanatory variables. The adjusted R2 of our regressions ranges between 0.40 - 0.70.                    42 Share of people RETIRED Share of people RETIRED Share of people WORKING Share of people WORKING Share of people RETIRED Share of people RETIRED Share of people WORKING Share of people WORKING 


































































































   42   43   44   45   46