A social accounting matrix (SAM) is an economy-wide consistent representation of the payments in an economy, linking production, primary factors, and institutions (the latter often split into households, government, and the rest of the world). In the words of Round (2003), “it is a comprehensive, flexible, and dis-aggregated framework which elaborates and articulates the generation of income by activities of production and the distribution and redistribution of income between social and institutional groups.” Most of the time, a SAM refers to the economy of a country for one year. It may be used to describe the structure of an economy and as a data input to economic models, most importantly CGE models. This paper documents a gendered Social Accounting Matrix (SAM) for South Korea for 2014: the steps followed in its construction, the data sources, and what the SAM says about the economic structure of this country (for brevity from now on referred to as Korea), including gender, time use, and the role of households in providing care and other services provided by households. The SAM that is presented is the key data input to a forthcoming analysis of gender and care in Korea’s economy based on GEM Care, a computable general equilibrium (CGE) model.

At an aggregate level, the value-added in the GDP economy is roughly 86 percent of the total (with agriculture, industry, services accounting for 2, 33, and 51 percent, respectively) and the non-GDP economy 14 percent. This information may be contrasted with aggregate time shares in terms of which, the non-GDP economy accounts for a much larger share, 40 percent. If leisure also were included, then it would on its own account for 50 percent of total time, 53 percent for men and 47 percent for women. The share of time allocated to household activities are drastically different for men and women, 15 and 57 percent, respectively. For women, non-care household services comprise the largest time using activity by a wide margin, followed by household childcare, trade services, and hotel and restaurant services. For men, manufacturing dominates, followed by trade services, household non-care services, and construction. For agriculture and industry, the lesser educated dominate all activities except electricity for both men and women. For services, the differences across sectors are drastic with some dominated by those with high education and others by those with low. These are few of the many descriptive findings from the gendered SAM developed by the authors.

This paper will be available December 2019.

This blog was authored by Binderiya Byambasuren, Kijong Kim & Hans Lofgren

A mother feeding a baby

Since the 1990s, unpaid work and care have garnered increasing academic attention, creating the emerging fields of the economics of unpaid work and the study of “the care economy.” Most of the earlier work was oriented toward microeconomics, focusing on issues such as the household division of labor, subsistence production in developing countries, the substitution between non-market and market goods and services in households, and the role of caring motivations in sectors of the labor market in developed and developing countries. Empirical work paralleled these theoretical efforts. Examples include measuring unpaid work via time-use surveys across the world, estimating the monetary value or opportunity cost of unpaid work, and linking care with the gender–wage gap and gendered job segregation. Macroeconomic models exploring gender and economic development were rare, and even fewer macro studies examined care and unpaid work.

Braunstein et al. (2011) contributed to this literature by embedding unpaid work and care in a structuralist macro model, illustrating how the social structures of production matter for economic outcomes. Braunstein and Tavani (2019) recast a simplified version this model in order to generate policy implications through numerical simulation. The authors go beyond the standard utilization-labor share framing in structuralist models in order to focus on more direct measures of gender inequality and the gender composition of the labor force and of employment. The resulting producer’s equilibrium – which describes the supply side of the model – always features a positive relationship between economic activity and gender wage equality. The goods market equilibrium, or the investment-savings curve, can either be care-led or inequality-led. The model results accommodate a variety of scenarios. A reduction in non-market care time can foster both aggregate demand and gender equality, but under some conditions it is also possible that there are trade-offs between economic activity and gender equality.

This paper will be available in December 2019.

Ipek IIkaracan Presenting Paper at Conference


İpek İlkkaracan, EMEL MEMIS, KIJONG KIM, TOM MASTERSON, and  Ajit Zacharias

Feminist economists have long emphasized the recognition, reduction and redistribution of unpaid care work (the so-called 3R strategy) as a primary policy intervention towards closing of the gender economic gaps. Investing in a social care infrastructure is an important component of the 3R strategy. Universal access to quality care services enables reduction of the unpaid work burden through its redistribution from the domestic sphere to the public sphere. Public investment and expenditure, however, is a question of fiscal policy at the helm of macroeconomists and policy-makers who are often gender blind and also adopt the mainstream approach of public expenditure constraint. Several waves of empirical research have examined the allocation of sectoral allocation of public expenditure through macro-micro simulations.

Yet until recently, studies have failed to explore an important outcome of investing in social care from a gender perspective, namely gendered patterns of time allocation. While investing in social care creates jobs for some women, enhances their access to employment and earnings, it also increases the requirements on their time through higher paid work hours. Simultaneous access to services alleviates the unpaid work burden of women with care dependent household members. The net welfare impact for different groups of women and men taking both time- and income-effects into consideration is an empirical question.

The forthcoming paper titled “Impact of Policy Interventions at Reduction and Redistribution of Unpaid Care Work On Employment Generation, Time- And Income-Poverty And Gender Gaps: A Macro-Micro Policy Simulation For Turkey” by Ilkkaracan, Kim, Masteron, Memis, and Zacharias uses an applied macro-micro simulation policy modelling approach to explore the gendered impact of increased public expenditures on social care service expansion on new employment and income generation, time allocation to paid versus unpaid work, time- and consumption poverty. An increase in social care spending triggers two types of effects at the household and individual level: It generates new jobs improving access of previously non-employed persons (predominantly women) to employment and income generation, but at the expense of increasing paid work time of job recipients. Simultaneously, improved access to childcare services reduces unpaid work time in the households with small children. The authors use a statistically matched data set from the 2015 Time Use Survey (TUS) and the Survey on Income and Living Conditions (SILC) in Turkey to assess the impacts on individual and household well-being, not only in terms of gains in employment and income, but also in terms of potential changes in household production responsibilities and time deficits.

The paper will be available December 2019

Cem Oyvat presenting paper at conference

According to the Global Gender Gap Index of the World Economic Forum (2018), South Korea is one of the lowest ranked countries in the world in terms of “Economic Participation and Opportunity” (124th out of 149 countries) as of 2018. The Global Gender Gap Index also shows that South Korea ranks 88th in terms of female labor force participation and 121st in terms of gender wage equality for similar work. The average wages of women in South Korea are on average 36.7% lower than average male wages (as of 2012, own calculations based on World Klems (2014) database). These statistics reflect that there is a significant economic gender gap in South Korea despite the fact that the country is now classified as a high-income economy. Moreover, the underdeveloped care infrastructure and reliance on unpaid care labor of women is posing serious demographic and social sustainability challenges in an aging society.

Existing research on the effects of public spending show a stronger positive effect of public spending in social care and education on female employment as well as total employment compared to public investment in physical infrastructure (Antonopoulos et al., 2010; Ilkkaracan, Kim and Kaya, 2015; Ilkkaracan and Kim, 2018; De Henau et al., 2016; Onaran, Oyvat and Fotopoulou, 2019a). These employment effects have further effects on the economy and the wellbeing of the society, as microeconomic studies across the board attest that a larger share of women’s income compared to that of men’s, is spent to satisfy the needs of the household (Blumberg, 1991; Antonopoulos et al, 2010; Pahl, 2000) and a possible increase in their income leads to increased spending on children’s education and wellbeing (Vogler and Pahl, 1994; Lundeberg et al. 1997; Cappellini, Marilli and Parsons, 2014). Onaran, Oyvat and Fotopoulou (2019a) confirm this finding at the macroeconomic level for the case of the UK. These consequently have further demand and supply side effects on output, productivity and employment (Onaran, Oyvat and Fotopoulou, 2019a; Seguino, 2017).

Oyvat & Onaran (2019) examines the short-run and medium-run impact of public spending in social infrastructure, defined as expenditure in education, childcare, health and social care on aggregate output and employment of men and women for the case of South Korea. They present a gendered Post-Kaleckian theoretical macroeconomic model. The authors estimate the macroeconomic effects of social expenditure using a vector autoregression (VAR) analysis for the period of 1970-2012. The results show that an increase in the public social infrastructure significantly increases the total non-agricultural output and employment in South Korea both the short and medium-run. Moreover, higher social infrastructure expenditure increases female employment more than male employment in the short-run and raises both male and female employment in the medium-run due to increasing aggregate output. Finally, estimates show that the South Korean economy is wage-led and gender equality-led in the short-run, hence overall equality-led, although the effects are economically small in comparison to the strong effects of increases social infrastructure spending, and become insignificant in the medium-run. The results indicate that sustainable equitable development and a substantial increase in employment requires a mix of both labor market and fiscal policies.

This paper will be available December 2019

Woman helping young child

The importance of public investment and adequate care policies for gender equality has come to the forefront of the policy agenda in recent years. The Sustainable Development Goals framework for the first time explicitly recognizes the unequal distribution of unpaid domestic work and care as main source of gender inequality. Target 5.4, in particular, draws attention to the role that public policies can play in providing infrastructure and social services to both reduce and redistribute domestic work and care., Gender-aware Computable General Equilibrium (CGE) models which account for different dimensions of gender inequality in the economy could be adopted to evaluate the gender distributional effects of these policies. Fontana et al. (2019) provide a review of this approach, highlighting the challenges therein as well as the ways forward.

CGE models are representation of the functioning of an entire economy drawing on detailed empirical data, and can be fairly disaggregated in terms of sectors, factors of production and household categories. Existing Gender-Aware CGE (CGGE) have been mostly used to simulate trade related policy changes and fall into two broad categories: those that use a gender-disaggregated approach and those that use a ‘two-systems’ approach. The first set of models disaggregate standard economic variables and categories such as production activities and labour factors to highlight female-intensive sectors.  However, unless these disaggregations are accompanied by behavioural rules that reflect the causes of unequal gender patterns, this approach does not generate plausible results

The second set of models are closer to a feminist economics approach as they include unpaid care work and its interaction with the market economy. However existing applications tend to model unpaid care work simply as a constraint to women’s participation in the paid market economy and are limited to comparative statics as opposed to dynamic analysis. There is therefore the need to develop a more nuanced approach that incorporates a range of labour market imperfections as well as   dynamic dimensions to better capture the gendered nature of both the market and the non-market economy. Such approach would   enable CGE modelers to better explore the distributional implications of alternative care related fiscal policy simulations. Fontana et al. (2019) also review CGE papers that are not gender-aware but incorporate labour market imperfections and dynamic processes that are relevant for gender analysis.

Using insights from these different bodies of CGE literature, they suggest several ways forward for modelling the gender distributional effects of public policies. First, modelers need to disaggregate their data sets not only by gender but also by other socio-economic characteristics to highlight how gender intersects with other sources of disadvantage and to reflect the specific gendered structure of a particular economy. For example, mothers of infants and elderly women face different challenges related to both work and care provision.  Second, modelers should consider different wage setting mechanisms and rules of operation of the labour market to better represent the causes of gender discrimination. For example, women in many countries tend to be crowded in a fewer sectors and occupations than men, including into care sectors that offer poor working conditions and pay.  Third, modelers need to move beyond modelling unpaid work as simply a constraint to labour market participation and emphasize its role in human capacity development in the long term. To this end, a dynamic approach would be especially useful.

The paper will be available December 2019