Educational Differentials in the Realization of the Intention to Have a Second Child in Italy.

Elspeth Graham, University of St Andrews
Francesca Fiori, University of Edinburgh
Francesca Rinesi, ISTAT Italian National Institute of Statistics

Increased female education is an important factor affecting contemporary fertility levels. However, less is known about how educational differentials in fertility relate to the (non)realization of fertility intentions. Although highly educated women do not usually express a preference for smaller families, they are often more likely to postpone the start of their reproductive career and to face higher opportunity costs of having a child. Some may therefore end up under-achieving or revising downward their fertility intentions.

This paper focuses on women in Italy who already had a first child and expressed their intention of having a second child in the future. We investigate whether the realization of their fertility intentions is associated with their educational level by asking:

  1. Do highly educated women differ from less educated women in the likelihood of realizing, postponing or revising their fertility intentions?
  2. To what extent can these differences be explained by other factors such as differences in age at first childbirth and in labour market engagement over time?

The study uses data from the Istat Sample Survey on Births 2005 (cross-sectional edition) and 2012 (longitudinal edition), which re-interviewed a sub-sample of the women who took part in the 2005 survey. Using multinomial regression, we model fertility ‘outcomes’ by 2012 for the sub-sample. Results show that highly educated women are more likely to realize, and less likely to revise downward, their fertility intentions compared to less educated women. These differences persist even when controlling for other biological and socio-economic factors conditioning their ability to realize fertility intentions. We also explore interactions between women’s educational level and these other biological and socio-economic factors. We conclude by discussing possible explanations for the findings, which provide additional insight into why Italy may be an exception within the EU-27.


Educational differentialsin the realization of the intention to have a second child in Italy.

Introduction

Increased female education is an important factoraffecting contemporary fertility levels (Ni Bhrolchain & Beaujouan, 2012). However,less is known about how educational differentials in fertility relate to the(non)realization of fertility intentions. Although highly educated women do notusually express a preference for smaller families, they are often more likelyto postpone the start of their reproductive career and to face higheropportunity costs of having a child. Some may therefore end up under-achievingor revising downward their fertility intentions, thus contributing to lowfertility in countries such as Italy.

This paper focuses on women in Italy who already had a first child and expressedthe intention of having a second child in the future. We investigate whether therealization of their fertility intentions is associated with their educationallevel. In particular, we ask:

a)     Do highly educated women differ from less educated women in thelikelihood of realizing, postponing or revising their fertility intentions?

b)      Towhat extent can these differences be explained by other factors such as differencesin age at first childbirth and in labour market engagement over time?

 

Research design, data and methods

The study uses data from the ISTAT Sample Survey on Births 2005 and2012 editions; especially from the 2012 longitudinal edition, whichre-interviewed a sub-sample of the women who took part in the 2005 survey.Women interviewed in 2005 all had a child in 2003, and at the time of theinterview were asked about their future fertility intentions. The 2012 survey thenre-interviewed the same women, updating their fertility histories and recordingtheir current fertility intentions.

We limit our attention to uniparous women who were interviewed in 2005and expressed the intention of having another child. We model their fertility‘outcomes’ by 2012, using a multinomial logistic regression model.

Our dependent variable has 4 categories:

a)      TheRealizers: women who had a second child by 2007;

b)      TheLate Realizers: women who had a second child between 2008 and 2012;

c)       ThePostponers: women who did not have a second child by 2012, but still intendedto;

d)      TheRevisers: women who did not have a second child, and had revised (downward)their fertility intentions by 2012.

The core explanatory variable ofour analysis summarizes women’s educational level in 2005 (Post-secondary,Secondary, Below secondary), at the time they initially expressed theirfertility intentions. The model controls for other factors regarded as relevantfor the (non)realization of fertility intentions. Thus, we include age at firstchildbirth, area of residence in 2005 and summary measures of partnership andlabour market histories between 2005 and 2012, and of the household’s perceivedfinancial situation.

Selected findings

Selected findings from themultivariate analysis (Figure 1) show the predicted probability of realizing,postponing or revising the intention of having the second child, according tomothers’ educational level.

Figure 1: Adjusted predictedprobabilities of achieving fertility intentions by 2012, by mothers’educational level in 2005.

 

It can be seen that highlyeducated women with post-secondary education have the highest probability ofrealizing their fertility intentions before 2007, and are the least likely topostpone or revise downwards their fertility intentions, if they still have notrealized them by 2012.

Concluding remarks

There are several possible explanationsas to why highly educated women are more likely to realize their fertilityintentions than women with lower levels of education. They tend to have betteremployment conditions and higher wages, and more often have partners with a higherincome; they may thus be more able to afford an additional birth despite higheropportunity costs. Further, they may have greater knowledge and awareness ofdeclining fecundity with age, which may encourage them to act on theirintentions sooner. Moreover, highly educated couples have lower dissolutionrates, which also favours fertility. Alternatively, highly educated women mightreport more realistic fertility intentions, if they anticipate better the biologicaland socio-economic factors that can influence their ability to have a child. Weexplore interactions between women’s educational level and these other factors toinform the discussion. Italy may be an exception in the EU-27 as it is the onlycountry where both mean actual and mean intended family size appear to behigher among highly educated women (Testa, 2014). Our finding that highlyeducated women in Italy are more likely than others to realize their fertilityintentions provides insight into why this might be so.

References

Ni Bhrolchain, M. and Beaujouan, E. (2012) Fertilitypostponement is largely due to rising educational enrolment?, PopulationStudies 66(3): 311-327

Testa, M.R. (2014) On the positive correlation betweeneducation and fertility intentions in Europe: individual and country-levelevidence, Advances in Life Course Research 21: 28-42

Presented in Session 1164: Fertility