Summary
Engelhardt and Prskawetz (2004) investigate why the cross-country correlation between total fertility rate (TFR) and female labour force participation (FLP) reversed from negative (pre-1980s) to positive (post-1980s) across 21 Organisation for Economic Co-operation and Development (OECD) countries. The core finding — consistent with concurrent work by Kögel (2004) and Engelhardt, Kögel, and Prskawetz (2004) — is that the within-country time series association between TFR and FLP never changed sign: it remained negative throughout. The sign reversal in the cross-sectional correlation is explained by unmeasured country-specific fixed effects and by country heterogeneity in the magnitude of the negative association. The paper then uses purely descriptive methods to identify which labour market, educational, demographic, and social policy variables explain why fertility declined most in low-FLP countries.
Key Claims
- The cross-country correlation between TFR and FLP turned from negative (
−0.8 in 1960) to positive (+0.4 by 1999); countries with highest FLP now have relatively higher fertility; the Mediterranean low-FLP countries (Italy, Spain, Greece) now combine lowest FLP with lowest TFR.
- Within-country time series associations between TFR and FLP did not change sign (shown by Engelhardt et al. 2004 error-correction models and Kögel 2004 pooled cross-section analysis). The cross-sectional reversal is an aggregation artifact driven by country heterogeneity in: (a) unmeasured country fixed effects, and (b) different magnitudes of the negative within-country TFR-FLP slope.
- Country grouping (dynamic, decade-by-decade): high FLP (>65%: Nordics, UK, USA, Canada), medium (50–65%: France, Germany, Japan, Australia), low (<50%: Mediterranean — Italy, Spain, Greece, Ireland, Belgium, Netherlands, Luxembourg). Netherlands moved from low to high FLP group over four decades (primarily part-time expansion).
- Factors consistent with overall TFR decline (Table 2 "Trend"): declining male labor participation, rising male/female unemployment, declining gender wage gap (rising opportunity cost), rising female education, declining marriage rates, rising divorce rates — all consistent with TFR decline via New Home Economics or proximate determinants.
- Factors explaining high TFR in low-FLP countries historically (Table 2 "Level by FLP"): higher female education in high FLP countries; higher first marriage rates in low FLP countries; lower divorce rates in low FLP countries — consistent with family norms and partnerships as fertility drivers in Mediterranean countries.
- Factors explaining more pronounced TFR decline in low-FLP countries (Table 2 "Trend × FLP"): (1) female unemployment rose most sharply in low FLP countries since late 1970s; (2) mean age at first birth increased more in low FLP countries since ~1985; (3) family allowances for first child declined in low FLP countries in the mid-1980s while remaining stable/rising in high FLP countries.
- Variables that cannot explain the pattern: proportion of part-time employment, female working hours, gross enrolment in pre-primary education — none show systematic variation across FLP groups consistent with fertility trends. In fact pre-primary enrolment was highest in low FLP countries (Mediterranean state subsidizes pre-primary but not flexible childcare).
- The "neither-work-nor-children" trap in low FLP Mediterranean countries: weak male economic position + traditional family norms + low family allowances + rising age at first birth compound to produce both low employment and lowest-low fertility.
Concepts Introduced or Extended
- Fertility-Employment Correlation — documents the cross-country sign reversal (negative to positive) and explains it as a cross-sectional artifact; identifies within-country TFR-FLP association as consistently negative; describes heterogeneity drivers.
- Nonmarital Fertility — marriage rate declines and age at first birth increases are proximate determinants linking FLP patterns to fertility.
- Demographic Transition — the paper characterises OECD countries' second demographic transition: below-replacement fertility combined with rising female employment, but the two are not universally positively linked within countries.
Entities Mentioned
Quotes
"The countries that now have the lowest levels of fertility are those with relatively low levels of female labour force participation while the countries with higher fertility levels tend to have relatively high female labour force participation rates."
"The role incompatibility hypothesis fails to explain the most central piece of the fertility-employment-puzzle: why are low female labour participation rates accompanied by low fertility rates, i.e. what causes women in some countries to make 'neither-work-nor-children' decisions?"
My Take
The paper's key contribution is separating the cross-sectional and within-country aspects of the TFR-FLP relationship — a distinction frequently confused in the policy literature. The descriptive Table 2 taxonomy is useful as a synthesis device. Limitations: purely descriptive, no causal identification, data coverage patchy across countries and time periods. The "neither-work-nor-children" characterization is evocative but not formally modeled. The role of ideational factors (gender norms, family structure norms) remains unmeasured. A follow-on paper (Engelhardt, Kögel, Prskawetz 2004 Population Studies) provides the formal econometric evidence using error-correction models.