Page 39 - IJPS-3-2

Do young children prohibit mothers from working in Ethiopia?

The older average age and the fewer number of urban children is probably due to the relatively higher educational
attainment of the urban women leading to the higher rates of contraceptive use (Table 2) and the mother’s delayed age at
first marriage/child bearing (Table 1).
Table 2 shows that, compared to the rural households, urban households have a higher proportion of female-headed
households, a lower proportion of households who received loan, a lower rate of maternal work participation, a lower
proportion of households with more than two children, and a higher proportion of households with members other than
parents who work for the household. While the higher proportion of female-headed households and the lower rate of
maternal work participation for urban relative to rural households are consistent with previous evidence, the lower rate of
loan receipt by urban compared to rural households is unexpected since urban households are expected to have better
access to the service given their proximity to credit facilities and the relatively capital-intensive nature of urban jobs.

Table 2. Demographic and economic characteristics of sample households (percentage)
Percent the event occurred
Variables
Full sample Urban sub-sample Rural sub-sample
Household head is female 19.7 29.3 10.1
Participant used contraceptives 51.5 74.4 28.6
Household received loan 43.5 36.2 51.5
Participant participated in productive work 47.7 38.8 56.5
Members other than parents participate in productive work 58.7 62.7 54.6
Members other than parents participate in non-productive work 65.2 67.3 63.1
First two siblings are the same sex 62.2 57.5 66.8
Households with more than two children 83.3 78.2 88.3
N 493 248 245
Source: Survey data (2010 and 2013).

3.2 Number of children and maternal employment status

Having described the characteristics of the study population, we now turn on to analyzing the effect of the number of
children on the maternal productive work participation, using the two step instrumental variable estimator of the ivprobit
model. Before that, however, we describe the maternal rate of work participation in relation to the number of children
(Table 3).

Table 3. Number of children and percentage of participants who participated in productive work prior to the survey
Full sample Urban Rural
No. of children
N % women working N % women working N % women working
2 82 34.3 53 31.2 29 42.6
3-4 148 46.8 86 55.1 62 37.9
5-6 149 42.1 63 21.8 86 55.1
7-8 76 60.7 32 36.2 44 77.5
9-10 30 58.2 14 27.3 16 86.2
≥11 8 100 --- 8 100
Total 493 47.7 248 38.8 245 56.5
Source: Survey data (2010 and 2013).
Table 3 depicts that the number of children and the work participation rate of the mother during the specified period
differ for the urban and the rural sub-samples. It shows that mothers who participated in productive work increase with
increase in the number of children for the rural sub-sample, and, generally, for the full sample. The situation is less
consistent for the urban sub-sample. However, the table clearly shows that the mother’s work participation is relatively
smaller for the urban sub-sample both on average and for each age group.
Table 4 and Table 5 present results of the ivprobit regressions (and for the exogenous probit model for comparison).
Table 4 shows the amount of variance explained for maternal work participation given the number of children and
control covariates. The p-value associated with the Wald χ statistic (ivprobit) is significant at p = 0.022 for the full
2
sample, suggesting that the model is well fit to the data overall, while this is not the case for the sub-samples. For the
exogenous probit model, however, the p-values are significant for the sub-samples as well.

International Journal of Population Studies | 2017, Volume 3, Issue 2 33