Modeling household fertility decisions in Greece

Bibliographic Data

ID12007495
AuthorsGeorge Hondroyiannis (0000-0002-4021-7517, Bank of Greece, Economic Research Department and Harokopio University, El. Venizelou 21, 102 50 Athens, Greece, corresponding author)
Year2004
Volume41
Issue3
Pages477-483
Publication date2004-06-21
Open AccessYes
TypeARTICLE
VenueThe Social Science Journal (JOURNAL)
Journal identifiersISSN: 0362-3319 • E-ISSN: 1873-5355
PublisherTaylor & Francis (PUBLISHER • GB)
DOI10.1016/j.soscij.2004.04.003
OpenAlexW2067544334
LanguageEN
Citations received5
References cited18

This paper develops an empirical model to investigate the main determinants of fertility in Greece employing cross section data. In the empirical analysis count models are employed. The results show that family income affects negatively fertility decisions suggesting that households prefer to provide quality services to their children than increasing the number of children. In addition, the results suggest that female characteristics, such as the level of education and the number of hours working in the labor market and the social status, affect inversely the fertility choice of the household. Finally, other female characteristics such as the health status affect positively the number of children. Overall, the empirical results provide support to the neoclassical theory of fertility. (1990), Ehrlich and Lui (1991), Sah (1997), and Cigno (1998) tried to expand Becker's original theoretical model.2 There are several empirical studies on household fertility decisions over the last years for industrialized and developing countries such as Carliner et al. (1980) for Canada, Nguyen-Dinh (1997) for Vietnam, Wang and Famoye (1997) for USA, Colombino (2000) for Italy, Kalwij (2000) for Netherlands, Melkersson and Rooth (2000) for Sweden.3 The ECHP survey is contacted by Eurostat and the data are collected by 'National Data Collection Units,' either national Statistical Institutes or research centers. The data is a large longitudinal data set that began in 1994 (first wave) with a sample of 60,500 nationally representative households with approximately 130,000 adults aged 16 years and over.4 EDUC3 is a dummy variable which is equal to 1 when the female completed the third level of education and zero otherwise, and EDUC2 is a dummy which is equal to 2 when the female completed the second level of education and zero otherwise.5 The value 1 represents the best health and 5 the worse. In particular, 1=very good health, 2=good health, 3=fair health, 4=bad health, and 5=very bad health. A negative sign in the estimated coefficients implies that an improvement in health increases the number of children.6 The table with the descriptive statistics is omitted due to space limitations. However, the table is available from the author upon request. The Jarque-Bera test statistic is equal to 2.82. Therefore, the null hypothesis of normal distribution is not rejected.7 Count models are employed when the dependent variable takes integer values that represent the number of event that occur.8 Due to space limitations the results of the estimated coefficients are omitted. However, all the results are available from the author upon request

Affect (linguistics · Demographic economics · Economics · Empirical research · Family planning · Fertility · Population · Research methodology · Sociology · Statistics · Total fertility rate · Demography · Family Dynamics and Relationships · Gender, Labor, and Family Dynamics · Global Health Care Issues

Tools

Open DOISci-Hub