Do imputed education histories provide satisfactory results in fertility analysis in the Western German context?



Zabel: Imputed education histories and fertility analysis in the western German context

would only be the case so long as it remains uncommon for people to reenter vocational
training after having children. Alternatively, in the future, educational trajectories might
change in such a way that people commonly gain additional vocational degrees later in
their life course even after having children. If reentry into vocational training thus
becomes independent of motherhood, false allocation of exposure time away from the
category ‘vocational degree’ into the category ‘in vocational training’ would then no
longer occur disproportionately often for the childless, and the risk of first birth for the
category ‘vocational degree’ would no longer be overestimated in imputed histories
using the last date the highest degree was attained. Given that scenario though, a
different problem would arise when using the last date the highest degree was attained
for the imputations. People who are actually not in vocational training would generally
be imputed to be in vocational training for long periods of time if they receive their last
degree comparatively late in the life course. Since birth risks are likely to be relatively
high in the periods when respondents are in reality not in vocational training but
represented to be in vocational training in the imputed histories, this would bias the
estimate of the risk of first birth upwards for the category ‘in vocational training.’

Thus it seems that it would be a better idea to use the first date the highest degree
was obtained than the last date if a very large proportion of people reenter vocational
training comparatively late in their life course. Using the first date would then involve
only a comparatively small amount of mis-imputation if durations of time spent in
vocational training after having received one’s highest degree for the first time are short
compared to the length of the gap between training spells. If, however, people
repeatedly reenter vocational training very often throughout their life course and the
amount of time spent in vocational training is altogether very large, but there are still
fertility differences between people in education and not in education, then estimation
results based on imputations using either the first or the last date the highest degree was
obtained are likely to be biased. In that case, it seems imputed histories could not be a
satisfactory substitute for original histories, and it would always be best to collect
information on complete education histories in surveys intended to provide data for
fertility analyses.

It seems that in the cohort analyzed in this study, not many people (only 4%)
obtained a university degree after having already obtained a vocational degree. If it
were more common to receive a university degree after a vocational degree and there
were large gaps between vocational training and university education spells, that would
also make imputation more problematic. The university degree, since it is the highest
degree, would always be the one for which the date is recorded, independent of whether
first or last highest degree dates are used for the imputations. It may be more common
for childless women with a vocational degree than mothers to take up university
education. Then, exposure time originating from the ‘vocational degree’ category would

152


http://www.demographic-research.org



More intriguing information

1. The name is absent
2. How Low Business Tax Rates Attract Multinational Headquarters: Municipality-Level Evidence from Germany
3. Corporate Taxation and Multinational Activity
4. The name is absent
5. The name is absent
6. PROTECTING CONTRACT GROWERS OF BROILER CHICKEN INDUSTRY
7. The name is absent
8. PACKAGING: A KEY ELEMENT IN ADDED VALUE
9. Assessing Economic Complexity with Input-Output Based Measures
10. The name is absent
11. The name is absent
12. Return Predictability and Stock Market Crashes in a Simple Rational Expectations Model
13. Testing Gribat´s Law Across Regions. Evidence from Spain.
14. The name is absent
15. The name is absent
16. The name is absent
17. Modelling the Effects of Public Support to Small Firms in the UK - Paradise Gained?
18. How Offshoring Can Affect the Industries’ Skill Composition
19. Concerns for Equity and the Optimal Co-Payments for Publicly Provided Health Care
20. The name is absent