Version classiqueVersion mobile

Who Cares?

 | 
Aatif Somji

5. Results and Analysis

Texte intégral

1The purpose of this chapter is to present findings from the primary data collection and conduct analyses of them in order to respond to the research question of whether unpaid care and domestic work is a significant constraint to female microenterprise development. In order to address this question in detail, it is necessary to assess: how much unpaid work is being done; what the self-reported constraints to enterprise development are; and how social norms on gender may be influencing the allocation of unpaid work. To maintain a practical, policy-driven focus, the research also considers possible solutions to address the distribution of unpaid care, both in absolute and relative terms.

2The chapter begins by describing the background characteristics of the female micro-entrepreneurs who form the study sample. Next, it turns to the time-use data to provide an indication of the allocation of time of these women, with a particular focus on unpaid work and how the division of this labour within the household specifically impacts them. The third section considers more subjective data on the self-reported constraints to enterprise development, starting broadly before narrowing in on unpaid work and on which specific interventions have the most time-saving potential in this regard. The final section considers social norms around gender, and how these could be contributing to the persistence of gender gaps in unpaid work, and by extension in returns to microenterprise. The chapter concludes with a discussion, summarising the key findings, elaborating on the extent to which unpaid work is a barrier to female enterprise development and suggesting what can be done to potentially mitigate this. Qualitative data from questionnaires and field notes are used throughout the chapter to supplement the quantitative findings and provide a richer understanding of the results.

5.1 Descriptive Statistics

3The selection of descriptive statistics provided in this section helps to give a basic indication of the background characteristics of the female micro-entrepreneurs sampled for the study. Respondents were selected out of a total population of female micro-entrepreneurs in Luwero District, Uganda, who were BRAC microcredit clients and had caring responsibilities for young children aged three to five years, estimated at 1,315. The total sample size of those who completed both the questionnaire and a valid time-use survey was 57 – roughly four per cent of the total target population.

Table 2: Descriptive Characteristics of Female Micro-Entrepreneur Respondents from Luwero District, Uganda

Characteristic

Mean

Standard Deviation

Minimum

Maximum

Age

37.47

10.37

20

61

Household size

6.04

2.40

1

12

Number of children

4.39

1.81

1

9

Rural

0.88

0.33

0

1

Sole business owner

0.90

0.31

0

1

No. of employees

0.39

0.84

0

4

Home-based business

0.32

0.47

0

1

Business income (USh/month)

369,825

461,958

30,000

3,000,000

Contribution to household income (USh/month)

168,509

114,322

10,000

600,000

Total household income (USh/month)

280,439

138,855

60,000

80,000

% contribution to household income

0.67

0.36

0.20

1

4Table 2 provides key background information on the women micro-entrepreneurs. The average age of the women was 37.5 years, though the spread was rather large due to both mothers and grandmothers being eligible respondents, as discussed in the Methodology chapter. The average household size was six, while the mean number of children was four. An overwhelming majority of respondents identified their household as being located in a rural area (88%).

Figure 2a and 2b. 2a - Marital Status of Respondents and Single; 2b -Cohabiting Respondents

Figure 2a and 2b. 2a - Marital Status of Respondents and Single; 2b -Cohabiting Respondents

5Considering the marital status of the sample, 44% were married at the time of asking, while approximately one third were separated or divorced. The remainder were roughly split between being unmarried and widowed. Exploring this data in more detail revealed that a sizeable proportion of those who reported being married were not actually living with their husband. Informal discussions with respondents suggested that this was largely due to polygamy. On the other hand, all those who reported being unmarried were in fact living with a partner. Thus, an alternative, and perhaps more useful way to summarise the data is to split the sample into cohabiting and single women, with the former consisting of those women – married or unmarried – who are living with a partner. This yields 22 cohabiting and 35 single women, 39% and 61% of the total sample respectively.

6Regarding education and literacy, 91% of the sample were able to read – almost all of whom could also write. The same percentage of respondents had attended at least primary school, with 46% having attended secondary school or above. These figures are broadly similar to the results of the most recent census data for Luwero District, which show a literacy rate for 18 to 30-year-olds of 90% and a female secondary school attendance rate of 41% (UBOS, 2018).

Table 3: Frequency Table of Business Types

Business Type

Frequency

Trading in primary foodstuffs

23

Agriculture (cultivating fruit, vegetables, etc.)

13

Restaurant, food stall

11

Basket-weaving

4

Hair and beauty

4

Selling charcoal

3

Rearing livestock

3

Drinks stall, bar

3

Tailoring

3

Petty shop vendor

3

7Moving on to characteristics related to the microenterprise, Table 3 shows the most frequently occurring business types. The modal business type was trading in primary foodstuffs, which refers to the selling, but not production, of basic food items such as fruit and vegetables. Over 40% of respondents were involved in this activity. The second most common business type was agriculture, which refers specifically to the production of crops and was often done in tandem with other food-related activities, such as the aforementioned trading in primary foodstuffs, or preparing food to sell at a stall or restaurant – the third most common business type. Overall, the three most common business activities among the sample were all are related to food and nutrition, lending credence to the observation made in the Case Study chapter, that Luwero District’s economic activities are largely agricultural in nature.

8Referring back to Table 2, almost all of the women (51) reported being the sole owner of their business, with the remainder having joint ownership over it. Approximately one third of the women reported that their main business was located at home, and the vast majority (77%) did not have any employees. The mean monthly income from the women’s businesses was 370,000 Ugandan shillings (USh), which roughly equates to 100 US dollars (April 2018) – though this figure varied greatly across respondents with a minimum value of 30,000 and a maximum of 3,000,000 Ugandan shillings. The average contribution made by the female business owners toward household income was less than half of their business income, suggesting that these women may be investing some of the latter back into their microenterprise. However, a caveat worth mentioning here is that informal conversations with the respondents suggested they did not consider expenses for things such as school fees to enter into the household income at all, and instead subtracted these expenses directly from their business income. Finally, total monthly household income was approximately 75 US dollars (April 2018), with the female micro-entrepreneurs on average contributing roughly two thirds towards this amount.

5.2 Time Allocation and Division of Unpaid Work

9The first part of this section uses the data from the time-use survey to provide a snapshot of how the fifty-seven women micro-entrepreneurs allocate their time, going into more granular detail on unpaid work activities. The second part then exploits data from the questionnaire to discuss the division of unpaid work and how this specifically impacts these women.

Table 4: Breakdown of Total Time Allocation

Time Classification

Mean

Standard Deviation

Minimum

Maximum

Paid Work

8.63

3.49

2

14.5

Unpaid Work

5.36

3.26

0

12

Non-work

2.60

1.63

0.25

7.5

Sleep

7.41

1.27

3.5

10.25

10Table 4 aggregates the individual time-use data to give an indication of how the women generally allocate their time within a 24-hour period on a regular working day. On average, they spend 8.6 hours on paid work, 5.4 hours on unpaid work and 2.6 hours on non-work activities, including eating, drinking, bathing, praying, socialising, using mass-media and resting. The remaining 7.4 hours are spent sleeping. In total, 14 hours of the day are spent dedicated to work of some form.

11The box plots in Figure 3 illustrate the distribution of hours across paid work, unpaid work, non-work activities and sleep. As can be observed visually, the variation in the number of hours spent on sleep and non-work activities are both relatively low. Conversely, paid work and unpaid work have a large variance. Combined, these findings suggest that differences between respondents are driven by these latter two variables.

Figure 3: Distribution of Total Time by Activity

Figure 3: Distribution of Total Time by Activity

12Table 5 disaggregates the data specifically on unpaid work to provide greater detail on these activities. Overall, 57% of total unpaid work carried out by the women is dedicated to caregiving services for family members – the recipients of which are generally their own children. One third of unpaid work is spent on domestic services, such as food preparation and cleaning of the house. The remaining ten per cent is allocated to the production of goods for own final use, which exclusively takes the form of subsistence farming within the sample. However, most respondents (79%) do not engage in this activity.

Table 5: Disaggregation of Unpaid Work Activities

Classification

Mean

Standard Deviation

Minimum

Maximum

Unpaid caregiving services for household and family members

3.04

2.69

0

8.5

Unpaid domestic services for household and family members

1.81

1.51

0

7.75

Production of goods for own final use

0.51

1.28

0

6

Unpaid volunteer, trainee and other unpaid work

0

0

0

0

13A qualitative assessment of the time-use surveys indicates that 36 out of the 57 women are simultaneously carrying out paid and unpaid work – 63% of the total sample. Many other women temporarily stop their paid work in order to do unpaid work. Overall, 81% of respondents are either pausing their paid work to carry out unpaid work, or conducting both simultaneously – suggesting that unpaid work may be constraining the ability of these women to dedicate their full time and attention to their business during working hours. This finding is in keeping with the hypothesis laid out at the end of the Literature Review, which raised the issue of diverted time, the difficulties is measuring this, and how it could be affecting the returns to capital for female micro-entrepreneurs.

Figure 4: Division of Unpaid Work between Women and their Partners

Figure 4: Division of Unpaid Work between Women and their Partners

14Moving from absolute to relative terms, Figure 4 shows the responses to a question on the division of unpaid work between the women and their partner. 77% of the female micro-entrepreneurs reported carrying out all of the unpaid work with respect to their partner, while 89% do either most or all of the caring and domestic responsibilities. However, this result is likely to be skewed by the high proportion of single women, as identified earlier in the chapter. When considering only those women who are cohabiting with a partner, however, 73% are still carrying out most or all unpaid responsibilities. Less than one quarter of cohabiting couples share these unremunerated tasks equally, with only a single reported instance of a man doing more unpaid work than his female micro-entrepreneur partner.

15The question on the broader division of unpaid labour between these women and their partner, whilst informative, remains fairly abstract. It could therefore potentially be more susceptible to self-reporting bias than more specific, detailed questions. To establish a greater degree of accuracy, the women were also asked about who most commonly carries out specific unpaid care and domestic activities, taking into account all household members. Partial results from this question, relating to the women themselves, are shown in Table 6.

Table 6: Total Unpaid Activities by Proportion Carried Out by the Women Respondents

Activity

Frequency

Proportion ‘self’

Care for ill or elderly

23

0.91

Childcare

56

0.88

Shopping

57

0.86

Cooking and food preparation

57

0.79

Washing clothes

57

0.70

Sweeping the floor

57

0.68

Subsistence

21

0.67

Collecting fuel

48

0.52

Cleaning the dishes

57

0.51

Fetching water

54

0.35

TOTAL

487

0.68

16The 57 women in the sample reported a total of 487 unpaid care and domestic activities, an average of more than eight discrete unpaid tasks per household – almost all of which are carried out on a daily basis. On aggregate, more than two thirds of all these activities are carried out by the female micro-entrepreneurs themselves. Moreover, 89% of all caregiving activities (for children, ill and elderly family members) are done by these women, suggesting that unpaid care is almost exclusively their responsibility. In sum, even when looking beyond partners to consider the entire household membership, 68% of all unpaid care and domestic tasks are carried out by the female business owners. Given that most of these tasks are daily activities, this finding adds further credibility to the time-use data suggesting that unpaid work is a significant constraint on the time of these women and could be impeding the development of their business.

5.3 Self-reported Constraints

17Having established how unpaid care and domestic responsibilities affect the women micro-entrepreneurs – both in absolute and relative terms – this section now turns to the more subjective, self-reported constraints to enterprise development. After establishing these broad constraints, particular focus is placed on unpaid work and the practical measures that can be taken to reduce this.

18Respondents were asked to rank the top three constraints they face to further developing their business, from a choice of ten options, plus other. A weighted score was then created for each constraint, according to the following formula, with the overall rankings shown in Table 7.

19Weighted Score = 3(Freq. Rank#1) + 2(Freq.Rank#2) + 1(Freq.Rank#3)

Table7: Self-reported Constraints to Enterprise Development

Constraint

Rank #1

Rank #2

Rank #3

Weighted Score

Access to Capital

45

9

2

155

Domestic Responsibilities

5

13

8

49

Access to Better Markets

3

11

9

40

Quality of Employees

2

6

11

29

Business Knowledge

1

6

4

19

Quality of Inputs

1

6

3

18

Control Over Own Income

0

2

12

16

Access to Savings Account

0

2

4

8

Other

0

1

2

4

20Access to capital was by far the biggest factor that these women believe would allow them to further develop their business, with almost 80% of respondents ranking it as their number one constraint. This finding was expected: informal conversations with BRAC research staff and primary investigators of the wider research project suggested that, based on their experiences, capital would habitually be cited as the major constraint. It also fits the narrative of the wider literature on microenterprise, which consistently sees lack of access to finance as the key factor constraining firm growth (e.g. Khandker et al 2013). This result also justifies the provision of microcredit to these women, whose capital constraint may have already been somewhat alleviated through the relaxing of credit constraints. Finally, it is worth considering the potential priming effect of the researcher being associated with BRAC: this may have encouraged respondents to overstate their capital constraint in the hope of receiving additional funds from the microfinance institution.

21While capital was overwhelmingly reported as the chief constraint to enterprise development, domestic responsibilities ranked second highest – despite no framing or prior information, suggesting that this was the focus of the research project. This surprising result adds considerable weight to the previous findings suggesting that unpaid work is a significant time constraint for these women and could be impeding the development of their business. Nine per cent of the sample ranked domestic responsibilities as their biggest constraint, while almost half of the women considered it one of their top three. Figure 5 shows how mean unpaid work varies according to the position of domestic responsibilities among the top three rankings. The more importance attributed to domestic responsibilities in constraining their business, the greater the amount of unpaid work carried out by the women on average. Most strikingly, those who ranked domestic responsibilities as their number one constraint spent on average almost seven hours on unpaid work per day, compared to the total sample mean of 5.4 hours. These findings suggest that many of those with significant unpaid care and domestic responsibilities are indeed aware of how these impair the development of their business.

Figure 5: Mean Unpaid Work by Domestic Responsibility Constraint

Figure 5: Mean Unpaid Work by Domestic Responsibility Constraint

22On the other hand, the mean unpaid work for those who did not rank domestic responsibilities as one of their three constraints was 5.2 hours – similar to those who ranked it as their second or third largest constraint. Therefore, while it is possible to show how average unpaid work can increase with the self-reported ranking, the data indicates that there are still many women carrying out an equivalent amount of unpaid work who do not consider it to be a constraint on their business. This result suggests three possibilities. For some women, their domestic responsibilities may simply not be constraining their business. Alternatively, they may not be aware of how these unremunerated activities impair their business. Finally, they may have other, more important constraints that have a greater effect on the development of their business.

23A final noteworthy observation from Table 7 is the third-highest ranked constraint: access to better markets. Beyond acting as a constraint on their time, it is feasible that women’s unpaid care and domestic responsibilities may also be limiting their choice-set with regards to business decisions, a theory outlined in the Literature Review chapter. Principal among these is geographical restriction, which can be the case when a woman’s business is based either within or very close to her home. Thus, a logical hypothesis that can be made here is that unpaid work influences women’s decisions on where to locate their business, which in turn limits their access to markets and therefore their ability to grow their business. While it is not possible to prove this causal link with the data collected, the argument is made more convincing by the fact that 100% of those who ranked access to markets as their principal constraint have home-based businesses, with a corresponding figure of 45% for those who ranked it as their second biggest constraint – almost 14 percentage points above the total sample mean.

24To sum up, the question on self-reported constraints to enterprise development demonstrates that access to capital was overwhelmingly the most common response. However, domestic responsibilities ranked second despite no indication given to respondents that this was the specific topic of study for the research. This suggests that unpaid work is indeed identified by many women micro-entrepreneurs as a key factor restricting the growth of their business. Finally, access to better markets ranked as the third-largest constraint overall, with suggestive evidence that this may be indirectly linked to unpaid care and domestic responsibilities through geographical constraints.

25The next question asked respondents how they felt about the amount of time they are able to dedicate to their business. The purpose of this question was to move beyond broad barriers to enterprise development, to focus specifically on the time constraint. The results are shown in Figure 6a. Somewhat surprisingly, given the results of the previous question, most of the women felt that the time they are able to dedicate to their business is just right. Only 13 of the 57 respondents reported that they would like to dedicate more time to their business – less than a quarter of the total sample.

26Qualitative comments from these women overwhelmingly cited care and domestic responsibilities related to children as the main reason why they were unable to spend more time on their business. For instance, several women reported leaving their workplace around lunch time to go home, prepare food and feed their children – before returning to the workplace later in the day. The quantitative data appears to validate these findings, with mean unpaid work by the women increasing with the extent to which they identify time as a constraint to their business (see Figure 6b).

Figure 6a Time Dedicated to Business

Figure 6a Time Dedicated to Business

Figure 6b: Mean Unpaid Work by Time Constraint

Figure 6b: Mean Unpaid Work by Time Constraint

27The final question on business constraints asked female micro-entrepreneurs how their business is affected specifically by their care and domestic responsibilities. Twenty-eight of the fifty-seven women reported that unpaid work impairs their business to some extent, while twenty-nine said that it does not affect their business – a roughly even split. What is puzzling about this result is its discrepancy with the previous one: only 13 respondents said that the time they are able to dedicate to their business is less than they would like, yet 28 reported their domestic responsibilities as impairing their business.

28Two possible hypotheses could explain this. Firstly, it could be the case that domestic responsibilities affect these women’s businesses in ways other than time. Alternatively, domestic responsibilities could indeed primarily affect their business through time, but this may only be made salient for many respondents when they are asked explicitly about these tasks. Qualitative comments from the women suggest the latter hypothesis is most likely. Similar to the responses from the previous question, those who reported their domestic responsibilities as impairing their business overwhelmingly stated the time devoted to caring and providing for their children as the main reason for this. Moreover, the time-use data appears to confirm this hypothesis. A t-test shows that, for those who reported domestic responsibilities to be impairing their business, the mean amount of unpaid work is 6.3 hours, while the corresponding figure for those who reported no effect on their business is 4.5 hours – a difference that is statistically significant at the five per cent level.

29Overall, the combination of the above three sets of results suggest that unpaid care and domestic work is identified by the female micro-entrepreneurs as a significant constraint to developing their business, directly and possibly also indirectly through the geographical restrictions these responsibilities place on these women. Roughly half of the sample reported their unpaid work to be impairing their business, with these women carrying out significantly more unpaid work than the rest of the sample. Finally, qualitative comments from the women suggest that caring and providing for their children constitutes the majority of this time.

30The final part of this section now turns to the practical steps that can be taken in order to reduce the unpaid care and domestic responsibilities of these female micro-entrepreneurs. The women were asked to rate how helpful a number of services would be in reducing their domestic responsibilities, from 0 (not helpful at all) to 4 (extremely helpful). The results are shown in the left panel of Table 8.

Table 8a: Rating of Services

Service

Freq.

Childcare

38

Household cleaning

9

Water point closer to home

5

Food preparation and cooking

4

Fuel source closer to home

1

Care for ill or elderly

0

Table 8b: Ranking of Service with Most Time-Saving Potential

Service

Rating

Childcare

3.02

Food preparation and cooking

2.07

Household cleaning

2.02

Water point closer to home

1.90

Fuel source closer to home

1.83

Care for ill or elderly

1.28

31Childcare received the highest ranking across all services, by quite a large margin. Similarly, when asked to select the one service with the most time-saving potential for them, two thirds of the women picked this option (see Table 8b). Given the qualitative responses from the previous questions, these results should not come as a surprise. Intriguingly, however, a t-test for equality of mean unpaid work between those whose eligible children do and do not attend day-care yields a p-value of 0.46, suggesting no statistically significant difference between them. So, despite childcare being seen as having the most time-saving potential, those women whose children attend day-care do not appear to be benefitting personally from this in terms of time saved on unpaid work.

32A more detailed analysis of the data sheds some light on why this may be the case. Firstly, 52 of the women send at least one of their eligible-age children to day-care, 91% of the total sample. Despite this, a majority of women still selected childcare as the service with the most time-saving potential. When asked why, the qualitative responses from the women follow a broad trend – and this is backed up by an in-depth qualitative review of the time-use surveys.

33In general, the women spend a significant amount of time in the morning preparing for their children: waking them up, clothing them and feeding them. A few then escort their children to their respective childcare facilities before going to the workplace. Once at the workplace, many begin preparing lunch while attending to customers. Most of the children attending day-care finish by 1pm and subsequently either return home or join their mothers at their place of work. Either way, the women are largely responsible for feeding them – diverting their attention away from customers at the workplace, or temporarily leaving in order to return home. After lunch, those who went home journey back to the workplace, while those who remained there balance their paid work with caring for their children. The women are generally responsible for preparing dinner for the household, which forces them to leave the workplace at a specific time – earlier than they may wish to. After returning home and bathing their children, the women serve dinner for their family – sometimes cleaning the dishes afterwards. They then spend some time caring for their children – talking with them or perhaps helping with homework – before finally preparing their children for bed.

34As is made clear in this vignette of the average working day for these women, the few hours their children spend at day-care do not appear to be having a transformative effect on the amount of unpaid work they carry out. Therefore, based on the responses to the initial question and the qualitative picture painted above, it seems that more hours of childcare services are key to enabling women micro-entrepreneurs to grow and develop their business.

35The next step in this analysis is therefore to explore how the length of time spent in day-care interacts with unpaid work. The questionnaire data indicated that 17 of the 57 women in the sample had at least one of their children in day-care for extended hours – defined as until at least 4pm. This facility was available in most of the day-care centres visited, at an extra cost. The total sample can therefore be split into three groups according to day-care status: those who do not send any of their eligible children to day-care facilities; those who send at least one child to regular day-care; and those who send at least one child to extended day-care. Figure 7 graphs the mean paid and unpaid work of the women according to this day-care status.

Figure 7. Paid and Unpaid work of the women

Figure 7. Paid and Unpaid work of the women

36There appears to be a positive relationship with paid work and a negative relationship with unpaid work carried out by the women micro-entrepreneurs, with respect to the amount of time their children spend in day-care. To test whether these associations are statistically significant, a regression analysis was carried out to estimate how the unpaid work carried out by the women varies in comparison with the extent of day-care their young children receive. Treating regular and extended day-care as two separate dummy variables, the estimation is as follows:

37Unpaidi = ß0 + ßi Regulari+ ß2 Extendedi + εi

38The results of this regression are shown in Panel A, Column 1 of the full set of regression results in Table 9. For regular day-care, the estimated coefficient is equal to -0.55 with a corresponding p-value of 0.72. The small coefficient and high p-value suggest there is little evidence of a relationship with unpaid work, compared to those whose children do not attend day-care. On the other hand, the coefficient on extended day-care is equal to -2.34, with a corresponding p-value of 0.16. While no statistically sound conclusions can be drawn from this result due to a p-value just above the 0.10 threshold, it nonetheless suggests a large negative relationship between extended day-care and unpaid work: those who send their children to extended day-care appear to spend over two hours less on unpaid work than those who do not send their children to day-care. Despite the small sample size, which makes statistical significance more difficult to establish, it can be argued that this result is still noteworthy – and indeed worth pursuing further.

Table 9: Regression Analysis Results

 

Unpaid Work

Paid

Work

Non-

work

Business Income

 

(1)

(2)

(3)

(4)

Panel A: Day-care Status

Regular Day-care

-0.55

1.04

-0.50

0.20

(1.53)

(1.65)

(0.92)

(0.38)

Extended Day-care

-2.34

2.64

-0.30

0.75*

 

(1.63)

(1.75)

(0.97)

(0.41)

Panel B: Extended Day-care Dummy

Extended Day-care

-1.86**

1.73*

0.14

0.57**

 

(0.92)

(0.99)

(0.55)

(0.23)

Panel C: Empowerment

Empowerment

-0.54*

 

(0.30)

 

 

 

Panel D: Cohabiting Status

Single

-1.59*

 

(0.87)

 

 

 

*** significant at the 1 per cent level, ** significant at the 5 per cent level, * significant at the 10 per cent level

39Given the small and highly insignificant coefficient on the relationship between unpaid work and regular day-care attendance, this group was combined with those whose children did not attend day-care to create a new dummy variable for extended day-care: taking the value 1 if at least one child attended day-care for extended hours and 0 otherwise. Doing this reduced the number of sub-groups of an already small sample, thereby increasing the statistical power of further regression results.

40Column 1 of Panel B shows the results of a regression of unpaid work on the extended day-care dummy, which suggest a strongly negative relationship between the two variables, significant at the five per cent level. Concretely, women sending at least one child to day-care for extended hours spend 1.86 hours less on unpaid work compared to those who do not.

41Given that these women appear to reduce their unpaid work, where do they reallocate this time? Columns 2 and 3 of Panel B show the results of similar regressions on paid work and non-work activities (including sleep) respectively. The regression of paid work on extended day-care yields a positive coefficient of 1.73 hours, significant at the ten per cent level. Interestingly, the increase in paid work associated with extended day-care is very similar to the reduction in unpaid work (1.86 hours). The regression of non-work activities on extended day-care, meanwhile, has a very small coefficient (0.14 hours) and an extremely large p-value (0.80), suggesting no relationship between these variables. Together, these results provide strong suggestive evidence that children’s attendance of day-care for extended hours enables women to reduce their unpaid work, with a corresponding increase in paid work.

42Continuing with this line of reasoning, an interesting avenue to explore is how the women’s business income varies with the day-care status of their children. A similar regression to the original one was carried out with the natural logarithm of monthly business income as the dependent variable, as shown in the following estimation:

43In(Business Incomei) = ß0 + ßi Regulari+ ß2 Extendedi + εi

44The results are shown in Column 4 of Panel A. The inclusion of the logarithmic function allows the coefficients to be interpreted as percentage changes. Once again, the coefficient for regular day-care is relatively small and the p-value large (0.60), implying that a relationship with business income is highly unlikely. By contrast, the coefficient for extended day-care is enormous – suggesting that sending children to day-care for extended hours is associated with roughly 75% more business income per month compared to not sending children to day-care at all. This result is significant at the ten per cent level.

45Following the logic of the previous set of results, the same regression was then run using the extended day-care dummy variable (see Panel B, Column 4). These results appear to confirm the strong positive relationship between extended day-care and business income: having at least one child attend day-care for extended hours is associated with 57% higher monthly business income, a result that is highly significant (p<0.02).

46Based on the combination of the previous two sets of results, a logical explanation is that sending children to day-care for an extended period of time reduces women’s unpaid care and domestic responsibilities, thereby allowing them to spend more time in paid work, which in turn increases their business income. Alternatively, the explanation could be one of reverse causality: higher business income allows women to put their children into day-care for longer. Both hypotheses are demonstrated graphically below. With the data available it is not possible to establish the direction of causality.

Figure 8: Alternative Hypotheses for the Relationship between Extended Day-Care and Business Income

Figure 8: Alternative Hypotheses for the Relationship between Extended Day-Care and Business Income

47To sum up, despite the majority of women already sending their children to day-care, they overwhelmingly stated that additional childcare services would be most beneficial to them in terms of reducing their unpaid care and domestic responsibilities. Further exploration of the data demonstrated that the time spent in day-care is usually only a few hours in the morning, with women having to adjust their work schedule to fit around this. There is no statistically significant relationship between unpaid work and sending children to regular day-care. On the other hand, those women who enrol their children in day-care for extended hours do significantly less unpaid work than the rest of the sample, while the relationship between extended day-care and business income is positive, large and significant. Together, this implies that extended childcare services may hold the key to reducing women’s unpaid care and domestic responsibilities and unleashing their economic potential in paid work.

5.4 Unpaid Work and Gender Norms

48The final section considers social norms around gender, and how these may be influencing the allocation of unpaid care and domestic responsibilities between males and females. Three sets of results are analysed: first, the detailed data on total unpaid activities is revisited to observe how many of these are carried out by females; next, a simple measure of empowerment is constructed to test if a relationship exists between decision-making power and unpaid work; finally, a set of attitudinal questions are used to estimate the opinions of women, men and the local community with regard to the role of men and women in unpaid and paid work respectively.

49Returning to the question of who most commonly carries out specific unpaid care and domestic tasks among all household members, Table 10 shows the proportion done by females.

Table 10: Total Unpaid Activities by Proportion Carried Out by Females

Activity

Frequency

% female

Cooking and food preparation

57

1

Washing clothes

57

1

Care for the ill or elderly

23

0.96

Sweeping the floor

57

0.95

Childcare

56

0.95

Cleaning the dishes

57

0.93

Shopping

57

0.90

Subsistence

21

0.86

Collecting fuel

48

0.67

Fetching water

54

0.65

TOTAL

487

0.89

50Beyond the previously established finding that women micro-entrepreneurs themselves carrying out 68% of all unpaid activities, these results show that fully 89% of all unpaid care and domestic responsibilities are carried out by females. Males, meanwhile, are commonly responsible for only 11% of total activities. The table ranks the activities according to the proportion of females responsible for doing them, which can be understood as a measure of their relative ‘feminineness’. Cooking and washing clothes are both unanimously done by females, while the percentages for sweeping the floor, cleaning the dishes and caring for children, the ill and the elderly are all comfortably above 90%. In general, it seems more acceptable for males to be responsible for fetching water or collecting fuel – though it should be added that females are still responsible for these activities in two thirds of instances. Both of these activities are associated with more intense physical exertion, for instance through carrying jerrycans of water or bags of charcoal, and therefore conform to a large extent with traditional gender stereotypes.

51A final insight to note regarding this result is around who these other females are. The female variable includes the women themselves, other female household members and maids. The low incidence of maids in general meant they only carry out three per cent of total activities. Informal discussions with the women micro-entrepreneurs established a common household composition of the women and their children, suggesting that the majority of residual unpaid work is therefore being carried out by their daughters.

52A series of seven questions were posed to the women to ascertain their level of decision-making power. The questions consisted of decisions about household expenditure on food, education and health; household savings; and expenditure, investment and savings of the women’s business income. A rudimentary measure of empowerment was then calculated, based on the number of choices over which the women had decision-making power. Empowerment therefore could take a value from 0 to 7, with a higher score suggesting greater empowerment. A hypothesis to test based on this data is of greater empowerment being associated with less time spent on unpaid work. The results of a simple regression analysis of these variables are shown in Panel C of Table 9.

53As hypothesised, there is a significant negative relationship between empowerment and unpaid work. More specifically, having decision-making power over one additional factor corresponds to 0.54 hours less unpaid work – significant at the ten per cent level. More detailed investigation of the data shows that the mean empowerment score is 6.39, with 44 of the 57 women attaining the maximum score of seven. This skew could be driven by the high number of single women in the sample (35), who are likely to make all the decisions about their business and household. To see whether this may be influencing the result, a t-test comparing mean empowerment of single and cohabiting women was run. Mean empowerment for these groups was 6.94 and 5.50 respectively, with the resulting p-value of 0.0001 strongly suggesting that there is a significant difference between them. As hypothesised, mean empowerment for single women is much greater – and extremely close to the maximum value of seven.

54Based on the assumption that empowerment is in fact being driven by single women, Panel D of Table 9 shows the results of regressing unpaid work against a dummy variable for being single. This yields a large, negative coefficient of -1.59, significant at the ten per cent level. Conceptually, it seems strange that single mothers spend 1.59 hours less on unpaid work than cohabiting women. However, given the apparent gender norms around the distribution of unpaid work, it seems plausible that the additional unpaid care and domestic responsibilities shouldered by women due to a cohabiting partner outweigh the latter’s own contributions to these activities. In other words, a cohabiting partner may add to the total unpaid workload of women rather than reducing it. Alternatively, it may be more socially acceptable for single mothers to establish informal arrangements within the community to receive help with their unpaid care and domestic responsibilities. Referring back to the alternative explanations to the Enterprise Household Model (Bernhardt et al., 2017) in the Literature Review, an interesting avenue for further research would be to explore the mechanism through which being single may be associated with higher returns to micro-enterprise, and whether this is being driven by changes in unpaid work.

55The final set of questions on gender norms asked the female respondents about two situations: the role of men in care and domestic work; and the role of women in income-generating activities. For each of these situations, the women were asked for their own opinion; what they thought was the view of their partner; and what they thought was the view of their local community. Women without a partner were asked about their most recent partner. The possible responses ranged from much less involvement to much more involvement, on a scale from 0 to 4 respectively. The distribution of results for the two sets of questions are shown in the box plots in Figure 9.

Figure 9a: Distribution of Attitudes towards Men's Role in Care & Domestic Activities.

Figure 9a: Distribution of Attitudes towards Men's Role in Care &amp; Domestic Activities.

Figure 9b: Distribution of Attitudes towards Women's Role in Income-Generating Activities

Figure 9b: Distribution of Attitudes towards Women's Role in Income-Generating Activities

56Regarding the role of men in unpaid care and domestic responsibilities, the women micro-entrepreneurs themselves were strongly in favour of greater involvement from them: more than 84% of respondents answered either 3 or 4 when asked their opinion. By contrast, there were mixed results when the women were asked about their partner’s view: 49% of women reported that their partners think men should be less involved, while the corresponding figure for more involvement was 42%. Finally, the women’s opinions on the beliefs of the local community were broadly positive, with 72% of respondents thinking that their local community wants men to be more involved in unpaid care and domestic responsibilities.

57Moving on to the role of women in income-generating activities, all of the female micro-entrepreneurs stated that women should be more involved to some degree, with almost 90% thinking that women should be much more involved. The women micro-entrepreneur’s beliefs about their partners’ opinions were also broadly positive, with 71% of them thinking that their partners wanted women to be more involved in paid work. Regarding the view of their local community, the results were overwhelmingly positive: 93% of respondents believed that their local community was in favour of women being more involved in income-generating activities.

58Two key findings can be drawn from these sets of results. The first is the large disparity between opinions on the role of women in paid work and the role of men in unpaid work. For the elicited opinions of all three actors, the mean score for the former was significantly higher than that of the latter – all valid at the one per cent level. This suggests that it is more socially acceptable for women to be more involved in income-generating activities than it is for men to be more involved in unpaid care and domestic activities. Informal conversations with the women suggested that this is the case: while it is normal for women to engage in paid work, there is a belief that it is still relatively uncommon for men to engage in unpaid work. Several women reported that others in their community would publicly shame men and their families for doing this work, while one woman – whose husband helped her with caring and domestic responsibilities – recounted being accused on multiple occasions of bewitching him. Finally, the implied reference point for this question is the status quo – which, according to the previous results, suggests an already unequal distribution of unpaid work between men and women. That women are encouraged to be more involved in paid work, without an equal amount of encouragement for men to be more involved in unpaid work, suggests that these inequalities will continue to persist over time, and perhaps even widen.

59The second key finding from these results is their apparent discrepancy with those at the beginning of this section. Despite the women micro-entrepreneurs overwhelmingly responding that men should be more involved in care and domestic activities, the majority of residual unpaid work appears to be done by their daughters. The fact that many of these women are single mothers suggests that they are actively allocating these responsibilities to their daughters rather than their sons, contributing to the reinforcement of prevailing gender norms across generations. When asked during informal conversations why they made their daughters do more unpaid work than their sons, most of the women had no response. Alarmingly, a few of them suggested that girls are responsible for domestic tasks while others claimed that boys are less likely to listen when asked to do unpaid care and domestic tasks. Overall, it appears that it is therefore necessary to shape both men and women’s attitudes and behaviour towards the role of males in unpaid work. Without achieving this, gender gaps in unpaid work are likely to persist – which may in turn contribute to the persistence of gender gaps in microenterprise returns.

Discussion

60The original research question of this thesis asks whether unpaid care and domestic work is a significant constraint to female microenterprise development. Deconstructing this question requires establishing how much unpaid work is being done by women micro-entrepreneurs; the extent to which they identify this as a constraint to their business development; and how the gendered distribution of unpaid work specifically affects females. The questionnaires, time-use surveys and informal discussions conducted with women micro-entrepreneurs in Luwero District, Uganda, provide an insight into the complex relationships between gender, unpaid work and enterprise development. Taking inspiration from the work of Diane Elson (2017), the findings and analysis of the research can be synthesised into three themes around unpaid work, each of which is explored in detail below.

Recognise

61The first theme focuses on the recognition of unpaid work. This can be understood in terms of accurately identifying the total amount of unpaid work being done by the women and acknowledging how this may act as a barrier to their enterprise development.

62The unpaid care and domestic responsibilities of the female micro-entrepreneurs are considerable. The time-diary method allows for a detailed appreciation of the allocation of time across different activities, with the data suggesting that these women dedicate an average of 5.4 hours to unpaid work on a regular, working day. Furthermore, over eighty per cent are either temporarily pausing their paid work to carry out unpaid work or doing both simultaneously. Either way, the time and attention of these women is being diverted away from their microenterprises during working hours, which is likely to have a negative impact on their business outcomes.

63Unpaid care and domestic responsibilities are also reported as a key constraint to enterprise development. Almost half of the women explicitly consider their unpaid work to be a significant barrier to growing their business – second only to capital on an aggregate scoring of constraints. Moreover, there is suggestive evidence that better access to markets, the third-biggest constraint according to the women, is linked to their business being home-based – a decision that may be influenced by expectations around unpaid care and domestic responsibilities.

64Finally, women who report that their caring and domestic responsibilities impair their business do significantly more unpaid work on average. From a policy perspective, this awareness is promising as it suggests a high take-up of programmes that seek to reduce unpaid work.

Reduce

65Having recognised how much unpaid work is being done by the women and how this may affect their businesses, reducing unpaid work then requires understanding how different physical and social infrastructure can alleviate this workload.

66Childcare is by far the service with the greatest time-saving potential according to the women. Most of them already send their young children to day-care for the morning, suggesting that this may not be sufficient in alleviating their time constraint. The provision of extended childcare facilities could therefore have a substantial effect on women’s allocation of time to unpaid work. Statistical analysis suggests that sending their young children to day-care for extended hours corresponds with almost two hours less unpaid work done by the women per day, which appears to be reallocated to paid work. Sending their children to extended day-care is also associated with 57% higher monthly business income for the women micro-entrepreneurs. Thus, it seems plausible that additional childcare services alleviate the burden of unpaid work for these women, freeing up time for them to dedicate to their business and thereby generate greater business income.

67Establishing a causal link between childcare, unpaid work and business outcomes is beyond the scope of this paper. However, the responses from the women coupled with the statistically significant findings from this small sample provide a strong justification for further investigation of this relationship. The forthcoming randomised control trial Child Care for Childhood and Business Development, to be carried out in Uganda by the wider research team, aims to establish the causal effect of extended day-care services for children on the business outcomes of their micro-entrepreneur mothers (NHH 2017). The eventual findings of this research are expected to make a substantial contribution to the nascent literature on the topic.

Redistribute

68While reducing total unpaid work may lead to welfare improvements, it does not change the fact that most of the residual unpaid work will be done by females. Transformative change therefore requires the redistribution of unpaid work, with a more equal sharing of these responsibilities between males and females.

69Unpaid care and domestic activities are overwhelmingly carried out by women and girls. In this study, 89% of all unremunerated tasks are done by them. Social norms on gender roles appear to explain why this may be the case. The role of men in unpaid work is consistently seen as less socially acceptable than that of women in paid work. Therefore, while women may be increasing their relative participation in paid work with respect to men, an inverse relationship does not appear to hold for unpaid work. These gender gaps in unpaid work are therefore likely to entrench gender gaps in paid work, such as in the business returns of micro-entrepreneurs.

70Social norms around gender and unpaid work appear to be embodied by both men and women. Despite wanting men to have a greater involvement in unpaid work, women allocate unpaid care and domestic tasks to their daughters far more than their sons. Complementary research conducted in Uganda by Oxfam shows that men and women overwhelmingly select girls as their first choice of helper for unpaid work, with both groups least likely to select boys from a choice of six possible helpers (Rost and Koissy-Kpein 2018). Together, these findings suggest that gender inequalities in unpaid work will persist without targeted interventions – for males and females – aimed at challenging these pervasive social norms.

Acheter

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search