CHILD HEALTH IN SOUTHEAST ASIA
Posted on : March 22, 2023Author : Nisha Malakar
INTRODUCTION
Countries in East and Southeast Asia have experienced an extraordinary pace of demographic and social change over the past five decades. Very high fertility rates in the 1960s and 1970s declined dramatically along with rapid socio-economic change in the region, due to rising income and education as well as the efficacy of family-planning programmes (Heuveline & Hirschman, 2015). At the same time, there has been little change in many other features of East and Southeast Asian families that are attributable to historical and cultural factors shaping family norms in the region (Raymo et al., 2015; Yeung et al., 2018). Consequently, it has been argued that trends in marriage and fertility reflect the tension between rapid social and economic changes on the one hand and limited change in family expectations and obligations on the other (Raymo et al., 2015), which could have been an incentive to introduce or intensify family policy interventions.
To get a first overview of similarities and diversity in family policy models in the countries of East and Southeast Asia, our goal is primarily descriptive. We provide a map of existing regional strategies for family support with respect to various instruments of family policy, embedded in the wider context of welfare regimes. We take a first look at the socio-economic, demographic and cultural factors which can help to explain the development of family policy programmes.
WELFARE REGIMES AND FAMILY POLICY IN EAST AND SOUTHEAST ASIA
The inclusion of family policy in the longstanding tradition of welfare state modelling in comparative welfare state research (Cutright, 1965; Esping-Andersen, 1990; Therborn, 1987; Titmus, 1974; Wilensky, 1974; Wilensky & Lebeaux, 1958) was the result of criticisms of the influential typology proposed by Esping-Andersen (1990). That is the extent to which social policy measures render individuals less dependent on formal wage-labour, and stratification, that is the extent to which social policy measures reproduce existing inequalities, hence disregarding unpaid work and the ways countries organize the provision of welfare through families.
The first wave of comparative social policy literature in the 1990s classified East Asian welfare states as exceptional cases that lie outside the three worlds of welfare capitalism (Esping-Andersen, 1990). In this strand of literature, the ‘Asian approach to social policy’ was characterised by a pervasive Confucian familism (Goodman et al., 1998; Jones, 1993), and a productivist approach to social policy (Holliday, 2000), i.e., state intervention is aimed at promoting economic productivity and growth rather than individual well-being.
Others contested the notion of welfare regimes aligning with regional blocks, exploring the balance between ‘productive’ and ‘protective’ dimensions of welfare state activity in a range of countries of Latin America, East Asia and Europe (Hudson & Kühner, 2012) and 29 countries in Asia and the Pacific (Kühner, 2015). Not only did the authors find little evidence that the commitment to higher ‘productive’ or ‘protective’ welfare was aligned along broad geographical units, they also found that the orientation towards ‘productive’.
CONCEPTUAL FRAMEWORK
As briefly mentioned in the introductory section of this paper, our operational definition of family policy includes transfers and services provided to families with dependent children. These include child benefits, direct and indirect family allowances, paid leaves and the provision of childcare services. These instruments can be mainly protective or mainly productive, depending on their level of (de)commodification, and they can have different implications for (de)familialisation.
Cash transfers reduce the direct and indirect costs of raising children (Hakim, 2011; Saraceno, 2011; Thévenon & Gauthier, 2011) and reduce the risk of child poverty (Barrientos & DeJong, 2006; Maldonado & Nieuwenhuis, 2015; Thévenon et al., 2018). As such, we consider them mainly a protective measure. However, depending on their design, they can be more or less decommodifying. If benefits are means-tested, they are targeted to low-income groups and are meant to relief poverty; thus, they do not eliminate the need for labour market incomes. Generous benefits may be an incentive for women to opt out of the labour market and take up care responsibilities, and can have familialising effects.
Paid leaves are an alternative for women to drop out of the labour market around the time of confinement. The length of paid leave has a small positive effect on female employment rates, as long as the total period is not any longer than approximately 2 years (Thévenon & Solaz, 2013). As paid leaves underpin the economic rationale of maintaining the functioning of the labour market, we consider maternity leave and parental leaves as mainly productive measures while the absence of care institutions may undermine women re-employment.
Childcare services and pre-schools help parents to reconcile work and family (Crompton & Lyonette, 2006; Stier et al., 2012) and create more employment opportunities for women, at the expense of family ties (Iversen et al., 2005). Early childcare education may also contribute to the long-term cognitive development of children, especially for those from disadvantaged groups (Drange & Havnes, 2019; Felfe et al., 2015).
Finally, cash transfers, leave policies and childcare services all affect timing and overall fertility decisions (Billingsley & Ferrarini, 2014) although their effect can be temporary where there is a lack of a comprehensive set of policies (Thévenon & Gauthier, 2011). Given this conceptual framework, we turn to the operationalisation of the instruments and analytical approach that we used to classify family policy regimes.
DATA AND METHODS
The data for this explorative study were collected from a variety of sources for the Welfare State Information System (WeSIS) of the University of Bremen. The sources include Social Security Programs Throughout the World, ILO Social Security Databases, UN World Population Policies Database, and others. In this paper, we use detailed information on various family policy measures in all 16 East and Southeast Asian countries that are included in major world databases: Cambodia, mainland China, East Timor, Hong Kong, Indonesia, Japan, Laos, Malaysia, Mongolia, Myanmar, Philippines, Singapore, South Korea, Taiwan, Thailand and Vietnam. Besides variables included in the database WeSIS, we assembled additional information for these countries and analysed survey data using World Value Survey (WVS).
Finally, we measured childcare provision by including the share of public expenditure in education dedicated to pre-primary institutions. We also included the gross enrolment rate in pre-primary institutions, and the share of enrolment in private pre-primary institutions, which should indicate actual access to public childcare.
We used the latest years available in the data sources (see Appendix, Table A1). The selection of the indicators was motivated by the objective to cover institutional (de jure) and de facto dimensions of the policies under study. However, the selection of optimal indicators was limited by data availability. The scarcity of data for all the countries in the sample also constrained the analysis to a single time point.
In the first step of our analysis, we used hierarchical cluster analysis to identify the countries with a similar set of policies. The aim of the chosen method was to distinguish clusters of countries that exhibit internal homogeneity in their type of family policy measures, such that cluster members are more similar to each
other than to countries in other clusters. The cluster analysis was run using Ward’s minimum variance method (Ward Jr., 1963), where the choice of a cluster for any given observation is made by the optimisation of the residual sum of square function. In other words, the objective of Ward’s algorithm is to minimise the within-cluster variance and maximise the between-cluster variance. As our dataset contains continuous and dichotomous indicators, we measured a pairwise Gower distance (Gower, 1971) which can be used to calculate distance between variables with discrete categorical and continuous numerical values. The resulting structure is visualised with a dendrogram (Figure 1).
FINDINGS
The result of the cluster analysis is visualised in Figure 1. The dendrogram displays distances or dissimilarities between the countries according to the combinations of family policy instruments described in the previous section. We opted for a four-cluster solution, cutting the dendrogram at distance 0.6. The first cluster includes Cambodia, Laos, East Timor and Myanmar (cluster A); the second cluster includes Singapore, Thailand and Vietnam (cluster B); the third cluster includes China, Philippines, Hong Kong, Malaysia and Indonesia (cluster C); and the fourth and final cluster encompasses Japan, Taiwan, Mongolia and South Korea (cluster D). Table 1 displays the average values of the indicators for each cluster. A more detailed description of the indicators by country is found in Appendix, Table 1.
TABLE 1. Indicators included in cluster analysis, mean values across clusters
Cluster A (Cambodia, Laos, East Timor and Myanmar): Maternity-oriented support
The profile of family policy in cluster A converges around benefits related to childbirth. All the countries in the cluster provide maternity leaves that fall somewhat below the minimum ILO standard of 14 weeks (98 days), with the exception of Laos that provides 105 days. Furthermore, all the countries in cluster A, with the exception of Cambodia, also provide short paternity leaves (Table A1).
Being employment-related, these provisions are hardly accessible to the large proportion of the population living in rural areas and outside formal employment. Expenditure levels for pre-primary education are low, at an average of 2.7% of total expenditure on education, and this is reflected in the low levels of enrolment in pre-primary education. The family policy model thus provides minimal social protection for families and, given the lack of care services, is implicitly familialising.
Cluster B (Singapore, Thailand and Vietnam): Employment-oriented support
Institutionally, the countries of cluster B provide fairly long maternity leave schemes that can favour the reintegration of mothers in the labour market after childbirth. Vietnam and Singapore provide paternity leave and Singapore provides parental leave (Tables 1) which is too short to be considered an instrument for sharing care responsibilities between partners.
The cluster offers low to medium social protection schemes. Two countries provide moderate redistribution among productive groups through contributory-based benefits, which are thus only accessible to children of parents in formal employment, and Vietnam provides means-tested benefits for poorer households. All countries in cluster B provide baby bonuses as a short-term economic buffer. Which is reflected in a high enrolment rate, particularly in Vietnam. Singapore also provides for flexible working arrangements. Thus, the family policy model in cluster B is moderately protective with a fairly strong productivist profile, which is rather defamilialising.
Cluster C (China, Philippines, Hong Kong, Malaysia and Indonesia)
Cluster C is very heterogeneous in terms of socio-economic structure and two different patterns can be observed. Indonesia, Malaysia and Philippines have comparably high fertility rates and even though they are formally secular states, they score very low in the disbelief index. In contrast, China and Hong Kong have very low fertility rates and are highly secularised. China has only recently started to proclaim the two-children policy, after decades of self-imposed low fertility levels due to the One-Child-Policy. Hong Kong, a British colony until 1997, may have been historically influenced by the British laissez-faire policy of minimal governmental interference in the economic affairs of individuals and society. Finally, the countries in this cluster score fairly low in choice and gender-egalitarian values, signalling that people are attached to traditional family models and gender roles and have little interest in defamilialisation.
Cluster D (Japan, Taiwan, Mongolia and South Korea)
Countries of cluster D provide a short maternity leave but long parental leaves (with the exception of Mongolia), maternity leave to self-employed women, flexible work arrangements, and benefits for the children (Table 1). Both universal and means-tested benefits can be found in this cluster. Japan has both contributory and means-tested benefits. Only Taiwan has a baby bonus. The family policy model in cluster D presents a medium-high level of protection, a moderate productivist and optional familialistic profiles.
To sum up, cluster A provides only basic instruments for the healthcare of a newborn and mother around the time of confinement, relegating the countries into an informal security regime (Abu Sharkh & Gough, 2010; Wood & Gough, 2006), which is implicitly familialistic (Leitner, 2003; Lohmann & Zagel, 2016). The countries of cluster C are economically more developed and provide, in addition to maternity leave, social assistance for children in poor families and a market for private childcare facilities. The productivist profile of the cluster is close to the facilitative profile identified by Holliday (2000), that is prioritising markets over social policy.
Cluster B provides instruments to facilitate transitions in and out of employment for working mothers as well as to compensate for the cost of raising children. Given the short maternity leave, high female employment and relatively high use of public childcare, the defamilialising profile of this cluster seems to be one of defamilialisation through public provision (Saraceno, 2016) or explicit defamilialisation (Lohmann & Zagel, 2016). Cluster D has relatively shorter maternity leave but provides the option of sharing care responsibilities between the partners. The productivist profile of this cluster is a mix of developmental-universalistic (Holliday, 2000) and social investment approach (Morel, Palier, & Palme, 2011), and its defamilialisation profile may be described as optional familialism (Lohmann & Zagel, 2016).
CONCLUSION
In this study, we examined governments’ strategies in East and Southeast Asia for providing support to families and reducing the costs of having children. We discussed the family policy models in relation to protective and productivist dimensions of East Asian welfare states and considered the implication for defamilialisation. We focused on child-related policies and three instruments of family policy: financial transfers, leave programmes and childcare services.
Our analysis provides insights into characteristic differences in family policies of the countries in East and Southeast Asia as well as hunches on possible drivers behind the different orientations. Our findings are also confirmative of other existing research. We find that the encompassing support model corroborates the argument that the productivist orientation of welfare regimes in East Asia is merging with a more inclusive model of developmentalism (Kwon, 2005). More generally, we find that family policy models cannot be deemed as simply protective or productivist, rather they exhibit the aspects of both. In line with Abrahamson (2017), we confirm that Confucian familialism is still widespread in the whole region. In particular, we note that such values are more resilient in individuals than in institutions, which show developments in favour of defamilialisation, particularly in the employment-oriented cluster.
A final substantive conclusion of this research is that the development of family policy in East and Southeast Asia appears not to conform to the familiar economically, culturally, and institutionally determined script known in Western welfare states, to which most of the literature refers to. Rather, family policy in East and Southeast Asia seems to be more fragmented and characterised by parallel interventions stemming from different provenances. In further research, it would be useful to assess individual countries and examine specific policies in greater detail than had been possible in this overview.
Nisha Malakar.
Intern, Asia in Global Affairs.
Leave a Reply