Short-term fertility intentions and their realisation in Poland

Previous studies identified large differences between countries in the extent to which childbearing intentions are realised. Failure to realise an intention to become a parent was found to be particularly common in the post-socialist countries. In this paper we examine whether similarly low rates of realisation of fertility intentions can be found in Poland. We use two waves of the Polish Generations and Gender Survey (GGS-PL), conducted in 2010/2011 and 2014/2015. We first describe fertility intentions of Polish women and men as declared at the survey’s first wave. Next, we examine whether the short-term childbearing intentions expressed at wave 1 were followed by an actual birth by the second round of the data collection. For the respondents who did not get a child between waves 1 and 2, we analyse the stability of their fertility plans. We find that approximately 35% of the respondents who at wave 1 intended to have a child in the next three years actually had one by wave 2. Both realisation and stability of fertility intentions varied markedly by gender and parity.


Introduction
Although the increasing availability of effective contraception enables couples to gain more and more control over their fertility, the childbearing decisions remain "imperfect" (Mencarini et al., 2015: 14). Not all family plans get realised and a substantial gap remains between the intended and actual family size (e.g., Liefbroer, 2009;Sobotka et al., 2015;Beaujouan and Berghammer, 2017). On the other hand, even in highly developed countries a non-negligible share of births are still declared as unintended or mistimed (Mosher et al., 2012).
In the European context, the most recent, cross-nationally comparative data on childbearing intentions and their realisation originate from the Generations and Gender Survey (GGS) (Vikat et al., 2007). These data have provided ample evidence on determinants of fertility intentions (e.g. Billari et al., 2009;Dommermuth et al., 2011;Klobas and Ajzen, 2015;Mönkediek and Bras, 2017;Mynarska and Styrc, 2014;Mynarska and Wróblewska, 2017;Neyer et al., 2013;Tanskanen and Rotkirch, 2014). The number of publications on the realisation of intentions is also growing (Dommermuth et al., 2015;Mencarini et al., 2015;Régnier-Loilier and Vignoli, 2011;Spéder and Kapitány, 2014). Importantly, the existing analyses have revealed marked cross-country differences in the extent to which an intention to have a child within the next three years (stated at the first survey wave) is followed by an actual birth by the second survey wave conducted three or four years after the first one. The realisation rates vary from 22% in Bulgaria, through around 30% in Georgia and Hungary to over 40% in France and Germany (Spéder and Kapitány, 2014); in Norway, where the data on childbearing behaviour were retrieved from administrative registers, they reached almost 60% (Dommermuth et al., 2015). A failure to realise short-term fertility intentions was argued to be "a key aspect of the post-communist fertility transition" (Spéder and Kapitány, 2014: 393).
Against this background, the Polish case has been only little explored so far. A panel survey conducted in 2007 and 2010 among four selected cohorts of young women living in two Polish cities reported realisation rates at 31% (Frątczak and Jeleń-Osiecka, 2015). Similar analyses carried out on a nationally representative sample of women and men are lacking.
In this paper, we provide the first overview of realisation of childbearing intentions based on a representative sample of Polish population of reproductive age. The existing Polish studies show that the level of fertility intentions as well as their determinants differ markedly between women and men, and between respondents of different parity: with zero, one and at least two children (Mynarska, 2011;Mynarska and Styrc, 2014). It appears that the process of fertility decision-making differs across these groups. Therefore, we analyse the pursuit of intentions also by gender and parity.

Data and methods
We use the first and second wave of the Polish Generations and Gender Survey -GGS-PL (Kotowska and Jóźwiak, 2011) conducted in the years 2010/2011 and 2014/2015, respectively. From the 8,404 respondents aged 18-45 who had taken part in the first wave, 4,760 (2,774 women and 1,986 men) were re-interviewed at wave 2. This number is equivalent to an attrition rate of 43%. While it is relatively high, it has not produced any substantial bias in the distribution of short-term and long-term fertility intentions (see also Brzozowska et al., 2018). Table 1 illustrates the basic sample characteristics weighted with post-stratification weights.
At both waves, the respondents were asked the same question about their shortterm fertility intentions: Do you intend to have a/another child in the next three years?
They could choose one out of four answers: definitely yes, probably yes, probably not, definitely not. Those who gave a negative answer (probably not or definitely not), were asked a second question, about their long-term fertility intentions: Supposing you do not have a/another child during the next three years, do you intend to have any (more) children at all? The second question allows us to establish whether a person intends to have any children at all, at any time in the future (long-term fertility intentions). For both questions, we dichotomised the answers, merging definitely and probably yes into one category, and definitely and probably not into another. Source: GGS-PL, wave 1 and wave 2, own computations, weighted data.
We conducted our analysis in two steps. First, in order to portray the spectrum of Poles' childbearing plans, we examined the distribution of short-term and long-term fertility intentions at wave 1. Second, we focused on short-term fertility intentions only and investigated their realisation and stability.
We classified the respondents into six groups that take into account the status of their short-term intentions at both waves and realisation of these intentions between the waves. We followed the approach taken by Spéder and Kapitány (2009) but added one new category ('newly intending'). The classification includes the following groups of respondents (see also Figure 1): • 'intentional parents': expressed a short-term childbearing intention at wave 1 and got a child by wave 2; • 'unintended parents': did not intend to have a child in the next three years at wave 1 (some of them planned this step in a longer time horizon) but got a child by wave 2; • 'postponers': did not get a child between the survey waves, but at both times expressed a short-term childbearing intention; • 'abandoners': stated an intention to have a child within the following three years at wave 1, did not get a child between the waves and abandoned their plan at wave 2; • 'consistently opposed': did not get a child between the waves and at both waves declared no short-term intention to have a child; • 'newly intending': did not plan to have a child at wave 1 and followed this plan, but at the second wave they changed their mind and intended to have a child within the next three years. All the distributions are constructed separately by gender and parity (childless, parents of one child, parents of two or more children), using the data weighted at wave one with the post-stratification weights. Note: The 'yes' answers include probably and definitely yes; the 'no' answers comprise probably and definitely not.
Source: own conceptualisation based on the GGS questionnaire.

Short-term and long-term childbearing intentions
Overall, at wave 1, around half of the respondents did not intend to have a(nother) child at all (Figure 2), 56% of women and 47% of men. As expected, these numbers varied greatly by parity: they were much lower among childless women (17%) and men (21%), and much higher among parents of two or more children (over 90% for both genders).
Among the parents intending further children, an overwhelming majority (fourfifths of one-child parents and two-thirds of the parents with at least two children) wanted to realise their plans within the following three years rather than later. By contrast, childless men wanting children planned to become fathers more often later than sooner (57% to 43%). Childless women planned their transition to motherhood within and beyond the three-year horizon about equally often. Among the parents with one child, men tended to plan the second one somewhat more often than women.

Realisation of short-term childbearing intentions
The overall realisation rate of short-term positive fertility intentions (i.e. intentions to have a child in the next three years) was around 35% (Figure 3). One-child parents were most successful in reaching their goal: 38% of women and 44% of men fulfilled their childbearing intentions. For the childless individuals and the parents of more than one child the figures were one-third and one-fifth, respectively, and were similar among women and men. Among the respondents who did not realise their shortterm intentions to have a child, the share of those abandoning this plan grew with parity. The difference between the childless respondents and one-child parents was modest (around 27% and 33%, respectively) but among the parents of two and more children the figures were substantially higher, reaching 64% for women and 57% for men. In contrast, the tendency to postpone fertility plans was inversely related to parity, starting from 10% and 17% among mothers and fathers, respectively, of two and more children and growing to around 40% among childless women and men.
Parity turned out crucial also for the realisation and stability of negative shortterm fertility intentions, i.e. intentions not to have a child in the next three years (Figure 4). The largest differences were seen in the 'newly intending' category: among the childless respondents it accounted for a meaningful share of 30%, but among the parents it was largely negligible. The 'consistently opposed' clearly dominated among the parents (80%-90%). Unintended births happened more often to the parents of one child than to those with larger families (11%-15% as opposed to 5%). Also a substantial proportion of the childless women experienced an unintended birth (13%; among men it was only 5%).  Finally, Figure 5 shows the complete classification of the respondents based on their childbearing intentions and actual reproductive outcomes between the survey waves. This overview allows us to highlight some of the patterns shown in the separate analyses of positive and negative intentions.
First, Polish women and men rarely plan to have another child if they already have more than one: almost 90% are 'consistently opposed' to increasing their family size. This number halves among the one-child parents and childless respondents. Second, the 'newly intending' are most often found among the childless respondents, while this category is virtually absent among the parents (between 2% and 5%). Notably, the number of 'unintended parents' varies relatively little by parity, especially among women (5%-7%). For men, the share is twice as high among one-child fathers (8%) than among other men. Lastly, women appear more determined than men to have their first child: the share of 'consistently opposed' is lowest among childless women (and by 10 percentage points lower than among childless men). They also intend to have their first child within the next three years more often than childless men. In contrast, men seem somewhat more inclined towards having a second child.

Discussion
This article provides an overview of gender-and parity-specific short-term fertility intentions and their realisation in Poland. Overall, one-third of the respondents realise their plan of having a child within the following three years. Thus, in the majority of cases, their declarations are not followed by actions. Those who do not plan any (further) children, are far more successful in carrying out their plan: over nine in ten did not change their parity by the second survey wave. However, quite a few of them revised their fertility intentions between wave 1 and 2, in particular, a notable share of childless individuals started to think about becoming parents by the second wave. This clearly illustrates an updating of fertility intentions related to life course developments: as one finishes education, finds a job or enters a cohabiting union, they start thinking about having a child (Heaton et al., 1999). At wave 1 a large share of childless individuals planned to become parents later than in the following three years. At wave 2, they updated their plan for the next three-year-time period. Importantly, also the births labelled as 'unintended' -that occurred between the waves mostly among one-child parents -should be interpreted in the view of life course changes in short-term intentions. These births were likely to result from the respondents' change of mind between the waves and not necessarily from an unwanted or mistimed pregnancy (see e.g. Williams et al., 1999).
Altogether, based on our findings, we can distinguish three distinct patterns of childbearing behaviour, related to the respondents' parity and with some notable gender differences. First, childless individuals generally plan to become parents (four-fifths have such an intention), but they do not rush their decision (one half would like to have a child within the next three years, whereas the other half prefers it to happen later). Unlike in the case of parents, for whom the gender differences in intentions to have a further child are modest, childless women wish for a child in the three-year horizon more often than childless men. This might result from a greater social pressure on parenthood put on women than on men (see findings for Bulgaria in Billari et al., 2009). Once the intentions are formed, however, childless women and men do not differ in their realisation, and they both postpone the birth of their children more often than parents do.
Second, over half of the one-child parents would like to have a second child. An overwhelming majority plan it sooner rather than later. They are also more successful in realising their intention than their childless counterparts. Both shortterm childbearing plans and their realisation are more common among men: they seem more determined than women to reach the family size of at least two children.
We may interpret this finding against men's lower involvement in childcare in Poland (Kocot-Górecka, 2015): as childcare usually burdens mothers much more than fathers, Polish women tend to be less enthusiastic about the idea of a second child. The dissatisfaction with the division of household and childcare tasks may also possibly result in a lower satisfaction with the partnership which has been shown to discourage women from planning a second child (Bernardi et al., 2013;Mynarska and Styrc, 2014).
Finally, parents of two or more children are generally highly reluctant to increase their family size, and the small group that plans to have further children is the least successful in fulfilling their intentions compared to the childless respondents and one-child parents. This reflects the relatively low share of families with three and more children in Poland compared to other European countries (Eurostat, 2018;OECD, 2011).
In the international context, the low realisation rate of fertility intentions in Poland is typical of the post-socialist countries of Central and Eastern Europe. It has been argued that this failure to fulfil childbearing intentions, observed in the post-socialist countries, "can essentially be traced back to the discrepancies between the value system and structural conditions of childbearing" (Spéder and Kapitány, 2014: 413). This explanation is likely to apply to the Polish case as well. Poles place a high value on family and children (Fokkema and Esveldt, 2008) and they wish to become parents (Mynarska, 2011). However, the economic and institutional constraints often play against their plans (Mynarska and Styrc, 2014). The unstable labour market, changes in regulations and family policies, difficulties in the housing market -they all might hamper the realisation of childbearing intentions. While our study offers first, descriptive insights into the topic, futher, more refined analyses are needed to verify what factors limit or foster the realisation of childbearing plans in the Polish context.

Short-term fertility intentions and their realisation in Poland Abstract
Previous studies identified large differences between countries in the extent to which childbearing intentions are realised. Failure to realise an intention to become a parent was found to be particularly common in the post-socialist countries. In this paper we examine whether similarly low rates of realisation of fertility intentions can be found in Poland. We use two waves of the Polish Generations and Gender Survey (GGS-PL), conducted in 2010/2011 and 2014/2015. We first describe fertility intentions of Polish women and men as declared at the survey's first wave. Next, we examine whether the short-term childbearing intentions expressed at wave 1 were followed by an actual birth by the second round of the data collection. For the respondents who did not get a child between waves 1 and 2, we analyse the stability of their fertility plans. We find that approximately 35% of the respondents who at wave 1 intended to have a child in the next three years actually had one by wave 2. Both realisation and stability of fertility intentions varied markedly by gender and parity.
Keywords: Generations and Gender Survey in Poland, fertility intentions and their stability, realisation of fertility intentions