Background Paper for LCR Regional Gender Study

World Bank

Time Use: Labor Market Participation and Care Demands on Women

Background paper for LCR Regional Gender Study

Case Study of Ecuador

1Introduction and objectives of the study

Feminist economics describes the relationship between labour situation of women and the accumulation market system as a conflictive process between production and reproduction spheres. As reproduction of the labour force involves a production process based on sexual division of labour, labour force in general, and women labour force supply particularly, are a consequence of this power process inside the household and the private arena, and express themselves as influenced variables into the market. Labour demand, on the other side, is based upon the capitalistic technical needs of production. Thereby, employment and distributive unbalances are the result of tensions between life conditions and accumulation (Picchio, 1994). The market, being an gendered institution, also reflects these and other generic constructions and reproduces inequalities and disadvantages of women, who enter and remain in the labour market in more vulnerable position than men, in most of the cases. Men’s labour condition is almost impermeable to familiar conditions, while they are a fundamental element for location of women in the market, and for the treatment employers give to both men and women. Thereby, occupational modalities (coming from labour policy or factual relationships) reflect the tension between participation and organization of the production and reproduction process. This tension has been solved by the market through labour flexibility (which is not always a full-rights regime), segregation and wage differentials. On the other side, invisibility of domestic care work as well as the less favorable entrance conditions of women to the labour market, have direct consequences in wage formation.

The objective of this paper is to analyze the relation between paid and unpaid labour for women in Ecuador, especially focusing on the influence of care demands on labour entrance and regimes, time dedicated to market labour and if market labour (and consequently income generating possibilities) can have an influence in distribution of care labour within the household. It must be said that, due to the endogenous nature of the relation between the two types of labour, statistical and econometric analysis will show a two way simultaneous relationship, more than pure impact. The ultimate interest of the study is to explore multidimensional characteristics of the interaction between private household and public market spheres, as it is a key piece of analysis in order to understand and improve women’s welfare situation.

2Time for care in Ecuador

Since the concept of care work is derived from “domestic” or “reproductive” work, it corresponds to a category not being considered productive (or work at all), from the traditional market-oriented economy point of view. Therefore, this type of work is included in what traditional labour theory calls “inactivity”[1]. This is one of the main reasons why information about domestic care activities have been ignored, overlooked and highly under-registered in labour and even household statistics.

Ecuador has one of the lowest inactivity rates among Latin American Countries. However, following the region’s tendencies, this percentage has been declining during the last 2 decades. In 1990, 30% of the working age population declared to be inactive, and this percentage felt to 26% in the first half of 2009, according to the ENEMDU.

Among inactive workers, population dedicated exclusively to domestic unpaid work (housekeepers) is almost completely female. In 1990 37% of female working population was dedicated to this activity. This participation has declined (more than declining of inactivity itself) rapidly, especially since the second half of 2005. Currently, 24% of female WAP are “housekeepers”. In the case of men, leaving the labour market is less frequent, but it appears from the data that 1999 crisis prevented some unemployed men to look for a job, and one part of this group dedicated their time to domestic work. This situation, however, was temporal.

Graph 1

Source: Employment and Unemployment National Surveys, various years

However, almost all women do unpaid domestic work (UDW). In general, according to 2007 National Time Use Survey, 95% or men and women over 15 years dedicates at least one hour per week to do this work. However, 85% women over 15 years dedicate more than 20 hours per week, while 80% of men dedicate less than 10 hours/week. The same source indicates that Ecuadorian society as a whole dedicates more hours to UDW than they do to market work. Measured by total working hours, UDW accounts for 52% of total work load (40% corresponding to women), and market work[2] accounts for the remaining 48% (18% corresponding to women). Therefore women contribute with 58% of total work load, 2 thirds of which are unpaid.

The “provisioning function” of domestic unpaid work can also be noticed if we consider the types of activities included in this complex work. According to the TUS, the most time/work demanding domestic activities are relating to food and cleaning (which includes laundry, which is a very important activity especially in rural area).

Table 1

Distribution of domestic Unpaid Work

Source: TUS, 2007

At an aggregate level, if we compare the working unpaid time dedicated to some sectors with the time in SNA activities/sectors, the results are interesting for the consideration of the UDW contribution to the economy. For example, if we consider the total unpaid domestic work dedicated to food activities and paid work in food industry and restaurants, UDW contribution is more than half of the resulting integrated sector. Likewise, total SNA hours dedicated to teaching, social protection, childcare and domestic paid work account for only one third of child care education and caring, if we add up child care domestic work.

Graph 2

Ecuador: Total working hours (CNS and unpaid non CNS) per sector

Source: TUS, 2007

This first description allows us to have an idea of the dimensions of the provisioning work, which is almost exclusively, a female responsibility.

At a micro level, we can see that this load is unequally distributed, not only between men and women. On a per week basis, women global workload (market work and UDW) is 1,4 times higher than men’s, and their domestic load is 3,2 times higher. This loadis higher for poor women, for a number of reasons (that will be described further). Graph 3 shows per week hours of UDW and market work on average per income quintiles, for men and women. We can clearly notice an inverse/direct relationship between market/domestic work (related to income generating capacity/incapacity) and poverty.

Graph 3

Source: Ecuador TUS 2007

The intuition behind this information is that higher work load is related to more children, lack of domestic services and infrastructure, less possibilities of access to own income, among others.

According to the TUS 2007, Ecuadorian women work on average 76.4 hours per week in both types of work, and men 60 hours. The difference is higher in rural areas, due especially to unpaid domestic work. In urban areas there is more intensity in market work, but unpaid market work is more generalized for rural women and men.

Table 2

Ecuador: Per week dedication to unpaid care domestic and market work by area

Source: Ecuador TUS 2007

Related to this, indigenous women and men present a higher work load than any other ethnic group in the country, and the highest gender gaps, as can be seen in graph 4.

Graph 4

Source: Ecuador TUS 2007

Another important dimension of UDW is temporal. Global work load is higher at ages between 25 and 55 years, when participation in labour market is higher also. But in general women, at all ages, have a higher load, even when both abandon the work force. Women dedicate more time to domestic work at ages between 25 and 30 years. Men, on the other side, have no significant variations in their dedication during their life cycle.

Graph 5

Source: Ecuador TUS 2007

Relating the family structure, it is not surprising to discover that the presence of children under 5 years old is related to a higher load of domestic care work for women, while the effect in men is not seen in domestic but in market work. However, having 2 or more children under five years represents a 30 hours per week increase in domestic female work (related to not having or having older children), but only a 5 hours/week increase in market work for men. Intuitively, this could mean needs coming from bigger families are absorbed in the first place by domestic work, due to market structures.

Additionally, women in nuclear (simple or expanded) households have a higher domestic load than women in other types of households, and women in nuclear simple household have the highest load of all types. Higher load are also present for married, as opposed to divorced or single women. This might give a clue about the relationship between sexual division of work and power relations within the household. Civil state marks high differences in workload, almost as much as the child number.

Graph 6

Source: Ecuador TUS 2007

Intensity of labour for women is concentrated in food, cleaning and childcare. Not surprisingly, the only activity where men dedicate more hours is management (includes reparations). However, load gaps are present in the rest of activities, more intensively in those related to food, cooking and serving. It is worth noticing that laundry activities are also a much generalized way of unskilled last resort work, which can be also a communitarian task, as well as making clothes, especially in rural areas (this is the case of older women).

Graph 7

Source: Ecuador TUS 2007

Poorer women tend to dedicate more time to child care and cleaning than richer ones.40% of poor households have more than 3 children under 5 years, and 30% are extended families. However, food related activities are not very different in intensity among social strata, with the exception of the richest quintile households. Care for seek and disabled persons (including old age) is not so relevant activity in general (despite health coverage is low even for high income households), but this could be due to under-registration of information about old age and people with disabilities.

3Time allocation for care and participation in the labour force

We can not understand market labour situation of women without having in mind their domestic work. As seen in the conceptual framework, sexual division of labour has direct and indirect consequences in the labour market, for women and for the market as a whole. Female participation on the labour force in Ecuador during last 30 years has been countercyclical and highly sensitive to changes in the economic environment. Historically, changes in this participation are more volatile than those of male workers, witha similar pattern being followed by unemployment (which is, on average 2 times higher for women). Graph 8 shows that, with the exception of 1999 crisis period, (especially female) labour supply has moved in this countercyclical unstable way. Unemployment and employment behaviour for women are equally more volatile than men’s.

Graph 8

Source: Vásconez, 2009

As it could be expected, recent crisis event has triggered the labour supply, for men and women, as well as unemployment, which in crisis tends to have a higher gender gap (2.5 times on average). From 80% at the end of 2007, male participation went up to 90% in March 2009; as for women, it moved from 52% to 62%.

Graph 9

Source: Vásconez, 2009

According to many studies for Latin America, the most important factors influencing labour force participation of women are related to their family relations and some human capital variables (especially when participating in formal sector). For Ecuador, a study based on data from LS Survey of 1999 showed that entering the informal sector is more related to labour context and family situation than to education or other individual characteristics. For all cases, though, child care services and access to family planning services are very important (Vásconez, 2002).

Relating entrance to the labour market, nearly 20% of women in working age (more than one million women) are “inactive” due to time constraints or family restrictions[3]. From earlier studies about women labour force participation in Ecuador, it can be seen that non participation is highly influenced by cultural aspects (being from the coastal region, where machismo phenomenon is notorious), localization (urban, rural), among others. According to the Employment and Unemployment (EUS) 2008 Survey, female inactivity substantially increases by having children, while male inactivity reduces for the same reason. Having children also increases the participation of women in part time jobs or underemployment, especially for female heads (22% of total). Table … shows that having children increases the proportion of inactive women (over women between 19 and 50 years) in around 10 points.

Table 3

Activity Condition by parenthood and sex

(Population between 18 and 50 years)

Source: EUS (2008)

Insertion to the labour market and its conditions are also related to family income, which is related (inversely) to domestic work, as seen. Non participation is directly related to poverty (women in poor households), and unemployment as well, showing lack of opportunities and incentives for poor women in the labour market.

Table 4

Market unemployment and inactivity by household income level

(Population between 18 and 50 years old)

Source: TUS 2007

This information gives a clue about determinants of participation in the case of poor women, which can be stopped by the presence of small children, lack of child care services (which is low in general for all households), infrastructure and domestic work facilities. Therefore, women with higher domestic work load are, at the same time, the poorest.

One of the most important direct consequences for women is their non-participation in the labour market. Even though there is not a complete trade –off between market labour and domestic labour time, there is a point where women have to stop working in the market in order to attend domestic duties, or not even offer her labour supply. On the other side, the provisioning task needs to be carried out, as seen before, in order to sustain reproduction and life itself; therefore, there is a basic amount of domestic work that has to be provided (and at the present time, it is provided by women in an unpaid way). We can see this “frontier” by correlating both types of work. Linear correlation is significatively negative, presenting a value of -0,198.

According to observed data, maximum work load is 110 hours per week (this value has been adjusted from the maximum remaining from total hours per week deducted self care, rest, seek periods, study and leisure). When developing a simple regression with alternate dependent/independent variable (market work and domestic work), we can reach to interesting results. We made this exercise for linear and quadratic functions, and linear functions gave better adjustments. The results are:

M = 54,983 – 0,684 D(1)

D = 67,901 – 0,682 M(2)

Where M is Market work and D is domestic unpaid work. Solving the system we can have the values of the highest combination of the two types of work, and the average values of substitution. Substitution rate is around 0,7, showing a relatively low slope, and also that there is not a direct trade off between the two types of work and also that, as we shall see later, there are a set of variables influencing labour decisions, apart from domestic work. The highest combination of both works is, on average, 54 hours of domestic work and 22 hours of market work, which corresponds to a part time job.

These values are very sensitive to income levels. For poor families and carers, labour supply is charged to domestic work, as for richer families, there is more “balance” between both types of work.

Since both types of work are sensible to the number of people within the household, I have adjusted time in per-capita levels. The results are interesting, since differences between poor and non-poor carers decrease a little, but gaps persist nearly unchanged.