Abstract
This article assesses 27 European parental leave policies regarding their compliance with an ideal leave policy model that best supports gender equality in the division of labour. Given the difficulties in defining such an ideal leave model, the article makes this assessment in two stages. Stage 1 exploits the most salient results of the empirical literature in order to define an ideal leave model that foresees the provision of 14 months of well-paid leave, where half of the leave is reserved for fathers. An ‘Equal Gender Division of Labour’ indicator is developed to assess the performance of different countries regarding their compliance with the ideal leave model. Stage 2 tests the sensitivity of the results with regard to three different scenarios that account for alternative assumptions about (1) the actual duration of leave that is supposed to be ‘ideal’, (2) the significance of providing leave only in combination with payments and (3) the importance of reserving some of the provided leave for fathers.
Introduction
This article assesses European parental leave policies regarding their compliance with an ideal set of leave policies that is assumed to best support gender equality in the division of labour. Driven by European Community law, there is a tendency towards the standardization of leave policies in Europe. Therefore, the assessment of leave policies with regard to one standardized ideal leave model lies at the centre of attention of policy analysis. However, the effects of different leave policies on mothers’ integration into the labour market and on fathers’ engagement in family work are only partially understood, and it is not clear how an ‘ideal’ leave model should be outlined. Therefore, the evaluation of leave policies with regard to one particular ‘ideal’ leave policy model has not been popular in previous literature.
However, given the recent developments in the literature studying the effects of leave policies on the gender division of labour, it is possible to draw several conclusions about the main characteristics of an ideal leave model (Dearing, 2015). First, there is convincing evidence that the duration of leave has a positive effect on the employment participation of mothers and their weekly working hours (Akgunduz and Plantenga, 2013; Genre et al., 2010; Misra et al., 2011; Pettit and Hook, 2005). However, these studies suggest that this effect diminishes with the duration of leave, indicating an inverted U-shaped relation. Therefore, very short and very long durations of leave only have a slightly positive effect on women’s employment participation and working hours, whereas moderate durations have a large positive effect. Second, at the same time, the duration of leave has been found to have negative consequences for other employment outcomes, such as wages, occupational segregation and job-related training for women (Akgunduz and Plantenga, 2013; Misra et al., 2011; Puhani and Sonderhof, 2011). In addition, research points to a widespread consensus that longer durations of paid leave delay a mother’s return to work after childbirth, implying that mothers actually use the extended right to paid leave if it is offered to them (Bergemann and Riphahn, 2011; Grunow et al., 2011; Lalive and Zweimüller, 2009; Ondrich et al., 2003; Ronsen and Sundström, 2002).
Concerning the effects of leave policies on fathers’ participation in family work, it is empirical literature from Sweden and Norway that provides strong evidence about a positive effect of the provision of leave reserved for fathers on their take-up of that leave (Duvander and Johansson, 2012; Ekberg et al., 2013; Kotsadam and Finseraas, 2011). Furthermore, a recent study by Pull and Vogt (2010) shows, using the example of the reform in Germany, that the introduction of a shorter and more generous leave benefit increased the number of fathers taking any leave. In addition, Lapuerta et al. (2011) emphasize the importance of the provision of financial support during leave. They show that the introduction of leave payments in several Spanish provinces has increased the number of fathers using leave.
Based on these results, I propose the evaluation of leave policies with regard to an ‘ideal’ leave policy model of 14 months of well-paid leave, half of it being reserved for fathers. This analysis will be based on a composite indicator – the equal gender division of labour (EGDL) indicator – that assesses the compliance of European countries’ parental leave policies to this ideal model. However, in order to account for blind spots in our understanding of how this ideal model should be outlined in detail, I propose a two-stage procedure for the assessment of leave policies. Stage 1 is the development of a baseline EGDL indicator. Stage 2 is a sensitivity analysis with regard to those assumptions that have to be made in the implementation of such an ideal model. This allows countries’ performance to be assessed under three different scenarios. The first scenario accounts for alternative choices with regard to the ‘ideal’ duration of leave set to 12, 14 or 16 months. The second scenario, proposes a mother-centred EGDL indicator assuming that gender equality in the division of labour is tackled best by the provision of a moderate duration of total leave in combination with a moderate duration of payments. The third scenario calculates a father-centred EGDL indicator that puts more weight on the performance of countries with respect to their provision of leave reserved for fathers.
My analysis builds on several articles engaging with the question of how parental leave policies and their support of a more gender-egalitarian society can be assessed within a systematic framework. The main concern with all of these contributions lies mainly with the deconstruction of the norms inherent in parental leave policies. However, they vary with regard to their methodological approaches. Exploiting the method of composite indicators, Gornick and Meyers (2003) assess the performance of 12 Organisation for Economic Co-operation and Development (OECD) countries according to their support of families through work–family reconciliation policies, where leave regulations are one of the measures examined. Besides their presentation of the generosity of leave across countries, they develop an ‘Index of Gender Equality in Family Leave Policy Designs’ and are, therefore, the first who explicitly account for the gendered structure in the systematic assessment of leave policies. Further elaborating this index, Ray et al. (2010) assess the generosity and gendered structure of leave policies in 21 high-income countries. A recent article by Javornik (2014) develops an index of state-familialism for eight post-socialist countries, measuring the degree to which a country supports women’s continuous employment and fathers’ engagement in family work based on policies on parental leave and childcare services.
From the perspective of policy typologies, Haas (2003) classifies leave regulations into ‘care policy models’ with respect to their duration of leave rights, the financial compensation offered during leave and the incentives for fathers to use leave. Based on similar leave characteristics, Wall (2007) suggests six main leave models and discusses them in relation to the availability of early childhood education and care services and female labour market participation. Using fuzzy-set analysis, Haas and Rostgaard (2011) assess leave policies according to their promotion of fathers’ leave take-up, in five Nordic countries. A recent article by Ciccia and Verloo (2012) also uses fuzzy-set analysis to study the conformity of European countries’ leave regulations to four theoretical ideal typical divisions of labour.
I aim to contribute to the previous literature with an evaluation of leave policies regarding their compliance with one particular ‘ideal’ leave model that best supports gender equality in the division of labour. For this purpose, I propose a methodological novelty that makes it possible to cope with some of the difficulties in the implementation of such an ideal model. In addition, the EGDL indicator accounts for non-linearity in the relations between leave policies and the gender division of labour, only barely considered by previous studies (see Ciccia and Verloo, 2012, as the only exception).
In my analysis, ‘gender equality in the division of labour’ is understood as a scenario, where, on average, men and women contribute equally to both the sphere of employment and family work. This concept has prominently been referred to as the ‘universal caregiver model’ by Fraser (1997) and ‘the dual-earner/dual-carer model’ by Crompton (1999) and Gornick and Meyers (2003). Given the actual distribution of work load, a ‘contribution towards gender equality in the division of labour’ would increase either the share of men’s investment in family work or the share of women’s engagement in employment. The design of the EGDL indicator implicitly assumes that leave policies that encourage fathers’ take-up also increase the share of family work they take on. Although it is reasonable to assume that there will be some positive effects, it is not clear to what extent the higher take-up of leave by fathers actually transfers into their overall engagement in family work over their life cycle.
Parental leave policies in Europe
Typically, there are three elements that make up a leave policy: maternity leave, paternity leave and parental leave. According to Moss (2013: 2ff), they can be described as follows. The first element, maternity leave, denotes an employment-protected leave of absence for mothers around the time of childbirth. This type of leave is generally understood to be a health and welfare measure, intended to protect the health of the mother and child. All European countries provide maternity leave regulations, mostly covered by earnings-related payments of between 60 and 100 percent of the mother’s previous income. The period of maternity leave usually amounts to a total of 14–20 weeks distributed before and after childbirth. The second element, paternity leave, denotes an employment-protected leave of absence for fathers after childbirth. This type of leave should enable fathers to spend time with their family right after childbirth. In contrast to maternity leave, paternity leave is only offered in some countries, such as the Netherlands, Portugal and Spain. It amounts to 2–15 days after childbirth and often comes with an income replacement rate similar to that applied for maternity leave. The third element, parental leave, mostly denotes employment-protected leaves of absence for employed parents – either the mother or father – which often follows the maternity leave period. Parental leave regulations vary widely across European countries – from 0 months in Switzerland to 4 years in Poland. Many countries provide payments during parental leave, either in the form of a generous income replacement payment for a shorter period or a less generous flat-rate benefit for a longer period. There are several options for building the policy objective of gender sharing into parental leave regulations. Most popular is the implementation of a ‘father’s quota’ that reserves some portion of leave for each of the parents or the provision of additional ‘bonus weeks’ of paid parental leave if parents share their leave entitlements.
The general use of the term ‘parental leave’ in the literature, however, is ambiguous – sometimes denoting an actual parental leave regulation in a narrow sense, while at other times applied as an umbrella term subsuming all three types of leave. In some countries, childcare leave is provided subsequent to parental leave, usually entailing an unpaid or poorly paid protected leave of absence from employment. In this article, the term ‘parental leave’, or simply ‘leave’, is used as an umbrella term and will refer to the sum of all these types of leave, including maternity, paternity, parental (in a narrow sense) and childcare leave. Furthermore, this article will distinguish between ‘total parental leave’ denoting the duration of leave, either paid or unpaid, and ‘paid parental leave’ denoting only that leave period that is supplemented financially.
Description of policy data
The calculation of the EGDL indicator is based on recent policy data from the year 2013 provided by the International Review of Leave Policies and Related Research (Moss, 2013). The review provides detailed policy data on parental leave legislation in 34 countries, 27 from Europe. Based on empirical results, three policy variables are of interest for the construction of the EGDL indicator: (a) the duration of total leave, (b) the duration of well-paid leave and (c) the share of well-paid leave that is explicitly reserved for the father. Table 1 shows the data on these three variables for the 27 European countries covered in the review.
Policy data on parental leave regulations in 27 European countries.
Source: Based on Moss (2013: 29f).
The ‘duration of total leave’ refers to the sum of paid and unpaid leave months. The ‘duration of well-paid leave’ considers only those leave months that are paid at a high flat rate of at least €1000 per month or two-thirds or more of earnings.The ‘share of well-paid leave that is reserved for the father’ refers to the portion of well-paid leave that is solely available to the father.
The data presented in Table 1 are mainly based on the aggregated data sheet of parental leave policies as provided by Moss (2013: 29ff) and refer to leave regulations as of April 2013. The ‘duration of total leave’ refers to the sum of paid and unpaid leave months. The ‘duration of well-paid leave’ considers only those leave months that are paid at a high flat rate of at least €1000 per month or two-thirds or more of earnings. The ‘share of well-paid leave that is reserved for the father’ refers to the portion of well-paid leave that is solely available to the father and encompasses paternity leave durations, ‘father’s quotas’ and ‘bonus months’.
Two adaptations have been made to the aggregated data table given by Moss (2013). First, the data given in Table 1 do not account for leave rights that are available to both parents within a certain time span twice. Second, in aggregating the data on the share of well-paid leave that is reserved for fathers, Moss (2013) does not consider the provision of bonus months when fathers take up their leave benefits. However, as these bonus months are very similar to ‘father’s quotas’, they too are considered as ‘leave reserved for fathers’ in Table 1.
When it comes to Moss’ (2013) calculation of the two variables on total and well-paid duration, all types of leave – maternity, paternity, parental and childcare – are subsumed into a broad concept of ‘leave’. It is, nonetheless, important to keep in mind that maternity and paternity leave periods are mostly not gender neutral, and their access is often restricted either to the mother or the father. This implies that caution should be used in interpreting the results, especially with respect to well-paid leave periods, which are frequently made up of longer maternity leave and shorter paternity leave periods.
Stage 1: development of the baseline EGDL indicator
First, I will discuss the theoretical framework that defines the relation between the three most crucial parental leave characteristics and gender equality in the division of labour, as suggested by the empirical literature. Figure 1 illustrates the theoretical assumptions about the non-linear relation between the duration of total leave and gender equality in the division of labour, as captured by an index with an inverted V-shape ranging from 0 to 1. The graph illustrates the assumption that the total duration of leave has an increasingly positive effect on mothers’ employment (and therefore on a more balanced gender division of labour), but only up to a certain point of ‘moderate duration’, after which the positive effect begins to diminish again. The index equals zero if the duration of total leave is zero or if it is very long (i.e., 49 months, which is the maximum value of total leave duration in Europe). The index equals one if the duration is ‘moderate’.
The design of the index assumes a non-linear, inverted V-shaped relation between the total duration of leave and the gender division of labour (GDL). The index assigns a value of 0 to countries with a minimum duration of 0 months of leave. With that point of reference, those countries with longer durations of total leave receive higher index values, up to the point of ‘moderate duration’ at which leave amounts to 14 months and the index value assigned is 1. After this point, the index values for countries with longer durations of leave decrease, down to a value of 0 for countries with a very long duration of 49 months.
According to this graph, the positive effect from the overall duration of leave is maximized at the ‘moderate duration’ of 14 months of leave. However, this is only an assumption, as the empirical evidence on the actual number of months that would denote a ‘moderate duration’ of leave is rather mixed and it is difficult to draw a conclusion. Therefore, in stage 2, I will consider different scenarios with regard to the assumption about the length of an ideal duration of leave and observe how the results change if the duration of total and well-paid leave is set to 12, 14 or 16 months.
With regard to the index based on the duration of well-paid leave, the empirical literature suggests that the duration of paid leave should not be too long, in order to ensure that mothers do not withdraw from the labour market for overly long periods. On the other hand, it has been shown that the provision of parental leave payments positively affects fathers’ take-up of leave. Thus, some moderate duration of paid parental leave best supports an EGDL. Based on these considerations, again, a non-linear inverted V-shape, as illustrated in Figure 1 for the duration of total leave, best represents the relation between the duration of well-paid leave and the gender division of labour. Again, the index equals zero if the duration of well-paid leave is zero or if it is very long (i.e. 25 months, which is the maximum value of well-paid leave duration in Europe). The index equals one if the duration of well-paid leave is ‘moderate’ (i.e. 14 months).
Considering index with regard to leave months reserved for the father, an ‘ideal’ leave model must acknowledge the findings testifying to the strong effect of the provision of leave reserved for fathers on their take-up of that leave, and thus I assume that such an ideal would reserve half of the duration of paid parental leave for fathers. Therefore, the index capturing the effect of leave reserved for fathers is assumed to follow a positive-sloped linear function represented by a straight line. The index equals a maximum value of one if half of the well-paid leave is reserved for fathers and equals zero if no leave is reserved for fathers.
In the next step, the baseline EGDL indicator is designed as an unweighted additive composite indicator computed from the three single indices and is calculated as follows
Figure 2 depicts the result for the baseline EGDL indicator, for the 27 European countries. I discuss the results with regard to five different groups of countries: those with ‘very high’ (between 0.8 and 1), ‘high’ (between 0.6 and 0.8), ‘medium’ (between 0.4 and 0.6), ‘low’ (between 0.2 and 0.4) and ‘very low’ (between 0 and 0.2) EGDL scores. The EGDL values indicate how well a particular leave policy complies with the ideal model of leave that provides 14 months of well-paid leave, half of it being reserved for fathers. There is no country that scores ‘very high’ on the EGDL indicator, implying that none of the countries provides these characteristics. It is Iceland that achieves the highest EGDL score of 0.76, providing 15 months of total leave, of which 9 months are well-paid and 3 are reserved explicitly for fathers. Sweden scores similarly well with an EGDL score of 0.74, providing 18 months of leave, of which 13.4 are well-paid, and also specifying a 2-month fathers’ quota and 10 days of paternity leave, both well-paid. Although Slovenia only provides 15 days of well-paid paternity leave, it is the third country in the high-scoring group. Its high performance can be attributed to the provision of moderate durations of total and well-paid leave, at 14.2 and 11.7 months.
Baseline EGDL indicator: Assessment with regard to an ‘ideal’ leave policy model of 14 months of well-paid leave, half of it being reserved for fathers.
In the group of medium-scoring countries, similarly, the performance of Denmark can be attributed mainly to the provision of moderate leave of 11.2 months, all of it well-paid. The medium EGDL scores of Croatia, Portugal, Finland, Germany and Norway mainly come from the provision of about 2 months of well-paid leave reserved for fathers. The design of these fathers-only policies, however, varies dramatically. Germany, Norway and Croatia provide fathers’ quotas of between 2 and 2.8 months. However, in Croatia, the approval of the two-month fathers’ quota is conditional on their prior use of 3 months of parental leave. Since a reform in 2013, Finland provides 9 weeks of well-paid paternity leave. Portugal is unique in reserving 30 days of maternity leave for fathers and, in addition, providing 20 days of paternity leave, of which 15 days are even obligatory to take. All countries in this cluster combine the provision of fathers-only leave with overly long durations of total leave (between 35.4 and 38 months). Ireland, Greece, Italy, the Netherlands and Luxembourg score in the low or medium range because, despite providing a reasonably moderate duration of total leave (between 13.9 and 20 months), they provide rather short durations of well-paid leave (between 1.9 and 6 months) and reserve no leave for fathers.
In the group of low-scoring countries, it is the EGDL composition of Poland that is exceptional. Poland performs well with regard to the provision of a moderate duration of 12.5 months of well-paid leave and it even specifies 2 weeks of paternity leave. The provision of an overly long duration of 48.5 months of total leave, however, places it among the low-scoring countries, as the design of the EGDL indicator reflects that this is a strong incentive for mothers to withdraw from the labour market for long periods. Another cluster of countries that score low on the EGDL indicator consists of France, Belgium, Spain and Estonia. These countries have low scores mainly due to their provision of overly long periods of total leave (between 35.7 and 36.5) in combination with short durations of payments. However, they all provide about 2 weeks of paternity leave and, therefore, still achieve EGDL scores of around 0.3, which can be interpreted as fulfilling ‘a third’ of the requirements of the ideal EGDL model. Compared to this cluster of low-scoring countries, Austria and the United Kingdom are much stronger with regard to their provision of total leave, at 24 months and 21 months. However, they provide only 1.8 months and 1.4 months of well-paid leave and no leave reserved for fathers.
Considering the EGDL results of very low-scoring countries, it first appears that the indicator compositions of Hungary, the Czech Republic, Switzerland and Russia have similar shapes. However, these countries actually have low scores for different reasons. The low scores based on the duration of total leave are due to long periods in Hungary, the Czech Republic and Russia, but a very short period of total leave in Switzerland (3.2 months). Whereas Hungary and the Czech Republic score low with regard to the duration of well-paid leave because they provide 24 months of payments, Switzerland and Russia do so because they provide only 3.2 months and 2.3 months. This result is driven by the inverted V-shaped design of the index on the duration of well-paid leave as discussed above. Lithuania, another very low-scoring country, provides 1 month of paid paternity leave. However, it receives an EGDL score of just 0.14 due to its provision of 36.9 months of total and 24.9 months of well-paid leave, thereby strongly incentivizing mothers to withdraw from the labour market for long periods. Slovakia brings up the rear with an EGDL score of 0.12 due to its provision of 3 years of total leave without any payments and no leave reserved for fathers.
Stage 2: sensitivity analyses of the EGDL indicator results
The ‘ideal’ duration of leave
The results of the baseline EGDL indicator rely on the assumption that the ideal EGDL model offers a ‘moderate’ duration of 14 months of leave that best supports gender equality in the division of labour. However, regarding the actual number of months that is referred to as ‘moderate duration’, empirical evidence is rather mixed and it is difficult to draw a conclusion. Whereas some studies find the positive effect to be maximized around 6–8 months (Akgunduz and Plantenga, 2013; Genre et al., 2010), others suggest an ideal duration of 18 months (Misra et al., 2011) or even around 3 years (Pettit and Hook, 2005). Therefore, it is important to validate the EGDL indicator results with different thresholds set to 12, 14 and 16 months of moderate and, therefore, ideal duration of leave (detailed results available from the author upon request).
However, the assumption of 12 months as the moderate duration of leave changes the countries’ EGDL scores and group composition only a little. It is only Denmark that moves up to the group of high-scoring countries as it is rewarded for the provision of 11.2 months of total and well-paid leave, which is now much closer to the ideal duration of 12 months. Similarly, applying the assumption of 16 months as the moderate duration of leave changes the baseline EGDL results only slightly. Slovenia drops to the group of medium-scoring countries, being ‘punished’ for the provision of 14.2 and 11.7 months of total and well-paid leave, which now lie further away from the ideal duration of 16 months. Similarly, Denmark’s score drops, as it is now punished for its provision of 11.2 months of total and well-paid leave.
The mother-centred EGDL indicator: combining moderate durations of total and well-paid leave
Leave policies are popular instruments in the battle against inequality in the division of labour because they can tackle inequalities on two fronts. On the one hand, they have the potential to balance the gender division in employment by fostering mothers’ integration into the labour market. On the other hand, they are a powerful tool for encouraging fathers to engage in family work. Those who focus on an equal division of work in the sphere of paid employment suggest that it is most important to provide a moderate duration of total leave in combination with substantial leave payments in order to support mothers’ involvement in paid employment. As there is no solid empirical justification on which to base this hypothesis, the following scenario checks for the sensitivity of the EGDL results to this assumption. Therefore, I assume a different mode of aggregation for the EGDL indicator: countries providing a balanced mixture of moderate durations of well-paid and total leave and which, therefore, assure that leave is provided in combination with payments, will be rewarded relative to those countries that score very well on one index and poorly on the other. Therefore, as opposed to the baseline scenario, countries cannot offset their weakness in one index with a strong performance in the other index. A mother-centred EGDL indicator is calculated, assuming that it is the combination of moderate durations of both total and well-paid leave that best supports gender equality in the division of labour
As can be seen in Figure 3, this assumption dramatically lowers the EGDL scores of all countries. This is due to the fact that all countries score less than one on at least one of the two indices – either with respect to the duration of total or of well-paid leave. Therefore, the product of these two indices will always be smaller than their sum. However, the mother-centred EGDL indicator penalizes some countries more than others. Those countries that perform well with regard to the combination of total and well-paid leave, such as Sweden with the provision of 18 months of total leave, 13 of it well-paid, score high on the index. Other countries that perform poorly on one of the two indices, such as Poland that provides a moderate duration of 15 months of well-paid leave in combination with 49 months of total leave – resulting in the provision of about 34 months of low or unpaid leave – score low on the index.
The mother-centred EGDL indicator: combining moderate durations of total and well-paid leave.
Therefore, besides Sweden, it is only Iceland that is still ranked as high-scoring on the mother-centred EGDL indicator due to its provision of approximately moderate durations of both total and well-paid leave. The mother-centred EGDL score of Slovenia, at 0.58, is only slightly lower than its baseline EGDL score of 0.64. This is due to the fact that Slovenia provides almost exactly moderate durations of both total and well-paid leave (with 14.2 and 11.7) and is therefore little affected by the new indicator. However, falling under the threshold of 0.6 it now joins the group of medium-scoring countries. Most formerly medium-scoring countries are downgraded to the group of low-scoring countries under the mother-centred EGDL assessment. Again forming one cluster, Norway, Germany, Portugal, Croatia and Finland get lower scores due to their provision of overly long durations of total leave (between 35.4 and 38.1 months).
Italy is the only formerly medium-scoring country that is downgraded to the group of very low-scoring countries. Under the baseline scenario, Italy scored high on the index due to its provision of 14.7 months of total leave. This performance is now compromised because this moderate duration of total leave is combined with only 3.7 months of well-paid leave. Belgium, Estonia, Spain and France drop from the group of low-scoring countries to the group of very low-scoring countries due to their overly long provision of total leave. However, they improve their performance relative to other counties such as Luxembourg and the Netherlands due to their provision of some leave reserved for fathers. Luxembourg and the Netherlands are among the countries most affected by the change to a mother-centred EGDL assessment. They drop from baseline EGDL scores of 0.38 and 0.39 to mother-centred EGDL scores of 0.09 and 0.13. Their formerly strong performance with regard to the provision of about moderate durations of total leave is now compromised by its combination with only 1.9 and 2.8 months of well-paid leave. Slovakia receives an EGDL score of zero as it provides no well-paid and no fathers-only leave.
The father-centred EGDL indicator: increasing the weight given to fathers-only leave
Another way to approach inequality in the division of labour would be to focus on encouraging a more pronounced engagement of fathers in family work by means of leave policies. There is consensus in the literature that well-paid leave reserved for fathers is a particularly effective way of increasing fathers’ take-up of leave (Bruning and Plantenga, 1999; Gornick and Meyers, 2003; Moss, 2008). To test the sensitivity of the EGDL results with regard to this assumption, a father-centred EGDL indicator is defined as follows
This alternative mode of aggregation alters the importance of the single components within the composite EGDL indicator. The baseline indicator calculates the performance of countries by weighting the single indicators equally. The father-centred EGDL changes this rationale by allowing the index with regard to the provision of leave reserved for fathers to dominate the assessment, determining one half of the composite indicator.
Figure 4 shows the results for the father-centred EGDL indicator. Almost all countries have lower scores than under the baseline scenario. This is due to the fact that the baseline EGDL scores of most countries are mainly determined by their strong performance with regard to their provision of total or well-paid leave, whose importance in the EGDL assessment is now cut back. Even Iceland and Sweden, the two high-scoring countries under this measure, have slightly lower scores under the father-centred EGDL indicator as their strong performance with regard to the provision of about moderate durations of total and well-paid leave now makes up only half of the EGDL indicator value. However, as most of this loss in scoring is compensated by the higher weight given to the provision of fathers-only leave, the two countries do still score highly under the father-centred EGDL assessment. Differently from the baseline scenario, the group of medium-scoring countries is now composed only of countries that support fathers’ use of leave through the provision of well-paid paternity leave or fathers’ quotas. Under the father-centred assessment, Croatia and Portugal are the only two countries whose scores increase compared to the baseline scenario, being rewarded for their reservation of a third of well-paid leave for fathers. The formerly medium-scoring countries of Italy, Greece and Ireland are now downgraded to the group of low-scoring countries, being punished for not reserving any leave for fathers. Estonia, Poland, Spain, France and Belgium do not experience a dramatic loss in their scores due to their provision of paternity leave periods. The Netherlands, Luxembourg, the United Kingdom and Austria all see their scores drop as they reserve no leave for fathers.
The father-centred EGDL indicator: increasing the weight given to fathers-only leave.
Robustness of EGDL results
The calculation of three different indicators allows us to check the sensitivity of the EGDL indicator results to different assumptions that cannot be justified empirically due to a lack of evidence. Figure 5 illustrates the robustness of the EGDL indicator results. As discussed above, the assumptions about shorter or longer durations of leave being ideal change very little with regard to the baseline EGDL results. It is only Denmark and Slovenia whose scores are volatile with regard to these assumptions, due to their approximately moderate durations of total and well-paid leave. Conversely, it can be seen that the mother- and father-centred EGDL assessments of leave policies change the scores of most countries a great deal. Here, the mother-centred EGDL assessment has the strongest effect. Applying this assumption in the EGDL assessment decreases the scores of all countries, with those countries that provide the most unbalanced combinations of total and well-paid leave affected the most. It is the high-scoring countries of Iceland, Sweden, Slovenia and Denmark whose scores barely drop under the mother-centred assessment. Similarly, all of the very low-scoring countries are robust to this change.
Robustness of EGDL results for all the different scenarios.
With regard to the father-centred EGDL assessment, the results are less volatile compared to the baseline scenario. Those countries that do not reserve any leave for fathers are most sensitive to this change in the indicator. Furthermore, it is again Slovenia and Denmark that are more sensitive in their EGDL scoring as their strong performance with regard to the provision of moderate durations of total and well-paid leave is compromised by the small weights given to these indices. The higher weight applied to the index of fathers’ leave does not offset this due to their rather short provisions of only 2 weeks of paternity leave. Again, it is the high-scoring Iceland and very low-scoring Slovakia, Lithuania, Hungary, the Czech Republic, Switzerland and Russia that are most robust to the change to the father-centred indicator. Croatia and Portugal even slightly increase their scores because they reserve a third of their leave for fathers only.
Conclusion
This article assesses the compliance of European countries’ parental leave policies to an ideal model of leave that would best support gender equality in the division of labour. In order to account for blind spots in our understanding of how this ideal model would best be implemented, I propose a two-stage procedure for assessing leave policies. Stage 1 develops a baseline EGDL indicator that is based on the most salient results from the empirical literature. Stage 2 provides a sensitivity analysis with regard to the assumptions that have to be made in the implementation of the ideal model. This allows us to assess the performance of countries under three different scenarios accounting for the ‘ideal’ duration of leave, a mother-centred approach and a father-centred approach towards tackling gender equality in the division of labour. As shown, some countries’ assessment scores are more robust to these different assumptions than others: Iceland and Sweden show high compliance with the ideal leave policy model regardless of which EGDL indicator is applied. Similarly, the group of very low-scoring countries is robust in their assessment of very low compliance with the ideal leave model. Conversely, the assessments of Slovenia and Denmark are highly volatile to almost all assumptions. In general, European leave policies are closer to ideal with regard to their provision of an ‘ideal’ duration of total or well-paid leave than with regard to the provision of an ‘ideal’ share of leave reserved for fathers. Similarly, as can be seen from the dramatic downgrading in the scores of all countries under the mother-centred EGDL indicator, European leave policies tend to provide either a moderate duration of total leave or a moderate duration of well-paid leave but not both in combination.
Footnotes
Acknowledgements
I thank Alyssa Schneebaum, Janet Gornick, Ulrike Schneider, Luise Gubitzer and Bernhard Blaschitz who provided helpful comments and suggestions in reviewing parts or drafts of this article. I have benefited greatly from the comments of the editor and two anonymous reviewers.
Funding
This article was partly developed during a DOC-team-fellowship of the Austrian Academy of Science at the Vienna Institute for Advanced Studies.