The Sustainable Development Goals were set by the United Nations General Assembly in 2015. These goals were intended to promote peace and prosperity for people and the planet, now and into the future.
Allocation of time within the household has been widely discussed since the paper by Becker (1965). Recent literature has incorporated time-saving appliance ownership into Becker's model. Greenwood et al. (2005) use this framework to explain increases in adult female labor force participation (LFP). They argue that appliances, such as washing machines and refrigerators, ‘liberated’ women from household work and allowed them to reallocate time to the workforce. Their model has been tested with data from the United States (Coen-Pirani et al., 2010) and China (Tewari and Wang, 2016), with both articles suggesting an increase in LFP among adult women who live in households that own time-saving durable goods. A similar relationship is found in Latin American countries (Cubas, 2016) and in Nigeria (Omotoso and Obembe, 2016, 2017). Cavalcanti and Tavares (2008) take advantage of a 24-year data set that contains the relative price index of home appliances for 17 Organisation for Economic Co-operation and Development (OECD) countries to examine the impact of home appliances on female LFP and find a larger increase in female LFP in countries that also experience larger drops in the relative price of home appliances, which supports findings in previous studies.
Kerr (2019) extends previous analyses by studying the relationship between these same time-saving appliances and child outcomes in China. She suggests that children reallocate time from household work to schooling when time-saving appliances are present in the household. Dao (2021) similarly examine the relationship between technological progress, housework, women in wage employment (nonagricultural sector), and daughters’ education in >100 developing countries. Again, their results suggest an increase in daughters’ schooling and an increase in female wage employment.
In this paper, we empirically test the hypotheses of the above researchers using microlevel data for Indian married women and children. This is the first paper to explore this question in the Indian context. India is unique in that, even with constant growth in overall LFP, India has experienced a decline in female LFP over the past 15 years. Female LFP reached its peak in 2005, with 32.17% of the female population aged ≥15 years participating. However, between 2005 and 2012, the country experienced a steady decrease in participation among females according to the International Labour Organization (ILO), achieving only a 20.02% female LFP rate in 2012. There was a slight increase from 2012 to 2017, but rates have fallen again in recent years.
This paper exploits cross-sectional variation in household appliance ownership to explore the effect of family investments in time-saving durable goods on (1) employment for married women and older children and (2) schooling for older children. We use microlevel data for the years 2004–2005 and 2011–2012 and address the concern of potential endogeneity of appliance ownership, which arises from unobserved family preferences that affect both appliance ownership and our outcomes of interest, by using an instrumental variable (IV) strategy. When estimating the effect of appliance ownership on a married woman's LFP, we follow Coen-Pirani et al. (2010) and instrument the ownership of a time-saving appliance by average ownership rate among
Our results indicate that the presence of a refrigerator in households reduces the probability of married women's employment by 10.2 percentage points. This result is counterintuitive and is the opposite of what previous literature has found in other countries. Thus, we further explore this question by examining the impact of these same appliances on two different types of employment: work Specifically, among married women, the probability of participating in animal care and family businesses (two lower paying employment opportunities) is found to be reduced and the probability of participating in agricultural labor and salaried positions (two higher paying employment opportunities) is increased. Detailed employment results are available in Tables A4 and A5 in the Appendix.
When analyzing children aged 12–18 years, we find no significant effect on enrollment or employment for children living in a household that owns a washing machine. However, our results suggest that children living in a household that owns a refrigerator experience a 15.3 percentage point increase in the probability of being enrolled and a 9.1 percentage point decrease in the probability of being employed.
These results imply that owning time-saving home appliances consequently avails family members the opportunity to reallocate their time from household chore work to other time-using activities, such as work outside the household for adult females and attainment of human capital in the form of schooling for children. The results presented in this paper suggest that one effective way to increase human capital development among children and increase female LFP among married women is to reduce household work obligations.
Appliance ownership in India has increased during recent decades, owing, in large measure, to rising disposable incomes, urbanization, increasing organized retail, easier financing options, growing demand, and increasing electrification (Ernst & Young, 2015). The proportion of Indian households owning a television increased from 47% in 2011 to 65% in 2018. Refrigerators are the second most prevalent household asset, followed by washing machines.
In this analysis, we follow Bowden and Offer (1994) by distinguishing between two types of household appliances: time-saving and time-using. Time-saving technologies embody appliances that reduce time required to perform household work. In the IHDS data, available time-saving appliances include washing machines, refrigerators, microwave ovens, electric cooking pots, and pressure cookers. Time-using goods, on the other hand, enhance the quality of discretionary time. Examples available in the IHDS data include air conditioners, air coolers, cameras, color televisions, and electric fans.
In this study, we focus on two large time-saving appliances: washing machines and refrigerators. Ownership of these appliances provides households the opportunity to substitute between capital and labor in home production. When a household purchases a time-saving appliance, household members have the opportunity to reallocate time to leisure, schooling, or market work. Previous literature on durable goods ownership has found that the presence of time-saving durable goods increases adult female LFP in the United States (Greenwood et al., 2005; Coen-Pirani et al., 2010), China (Tewari and Wang, 2016), Latin American countries (Cubas, 2016), Nigeria (Omotoso and Obembe, 2016, 2017), and OECD countries (Cavalcanti and Tavares, 2008). These same appliances are found to increase school enrollment and decrease child LFP in China (Kerr, 2019). Time-conserving appliances provide women and children with more efficient ways to perform domestic tasks, making them more productive in the household, and thus increasing their intrafamily bargaining power.
This paper adds to the literature by studying the effect of household appliances on Indian adult females and children. With the drastic changes in appliance ownership occurring in India, as described above, and the cultural differences between India and the countries studied in previous research, exploring the relationship between appliance ownership and female and child outcomes within the Indian context is an important contribution to current literature.
As described above, appliance ownership has been found to be associated with increases in female LFP in multiple developed and developing countries. While prevalence in these same appliances has increased in India, India's female LFP has been declining while remaining visibly low. The ILO ranked India as the country with the 28th lowest female LFP in 2019, as female LFP declined from 31% in 1990 to 23.4% in 2019. These are estimates of the participation rate of the female labor force (percentage of the female population aged ≥15 years).
One hypothesis as to why female LFP in India has declined even with extreme economic growth is that the Hindu culture and the caste system create a need for differential preferences for women. While Chinese women have increased their LFP with increased technological advancements, Indian women face different obstacles. India's average household size is 1.2 persons larger than China's, suggesting that family resources may be more diluted and require a larger time commitment at home for women.
Understanding the underlying trends in female LFP is important to those interested in women's well-being. Beyond women's contribution to growth, being economically active affects women's progression toward economic independence, bargaining power, and their children's overall well-being (Klasen and Pieters, 2015; Mammen and Paxson, 2000). With India's decrease in female LFP, it is important to explore whether the appliances found to improve female outcomes in other countries also improve female outcomes in a country that is currently experiencing a consistent decline in female LFP. This study thus helps understand the relationship between appliance ownership and female LFP in the Indian context.
The education system in India, also known as the “10 plus 2” system, consists of 4–5 years of primary education, 3–4 years of middle education, 2–3 years of secondary education, and 2 years of senior secondary education. Children begin primary education by the age of 5 years.
According to Government of India's Ministry of Human Resource Development, the total gross enrollment ratio (GER) at the primary level increased from 109.4 in 2005 to 116 in 2012. Gross Enrollment Ratio (GER) is defined as the total student enrollment in a given level of education, regardless of age, expressed as percentage of the corresponding eligible official age-group population in a given school year. This allows the GER to be >100%. India's Right to Education Act makes schooling free and compulsory for children aged 6–14 years, which partially explains why GER drops during middle education.
Kerr (2019) indicates that the presence of time-saving household appliances alleviates children from household work and provides them the opportunity to reallocate time to schooling in China. With the low secondary and senior secondary enrollment rates present in India, it is useful to explore whether these same appliances can help initiate similar changes in school enrollment rates.
The data used in this analysis come from the IHDS. This survey was conducted by the University of Maryland and the National Council of Applied Economic Research, New Delhi. The IHDS is a nationally representative, two-wave longitudinal data set consisting of a first round of interviews completed in 2004–2005 and a second round of interviews, mostly reinterviews of households from the first wave, completed in 2011–2012. The first wave contains data on 41,554 households in 1,503 villages and 971 urban neighborhoods across India, and the second wave contains data on 42,152 households in the same villages and urban neighborhoods as the first wave. For a full description of the survey design, see:
Our first sample consists of 52,046 married women of working age, with 73,916 observations over the 2 years of data. Table 1 presents the descriptive statistics for this sample, partitioned by household ownership of a washing machine and a refrigerator. Less than 10% of the sample lives in a household that owns a washing machine. Refrigerator ownership is more prevalent, with nearly 30% of the sample living in a household that owns a refrigerator. The summary statistics are similar regardless of whether we partition the sample by washing machine or refrigerator ownership. Thus, we discuss general differences between SC-ST refers to scheduled caste–scheduled tribe, the historically underprivileged sections of the Indian society.
Summary statistics by ownership of appliance: married women aged 15–59 years
Age, years | 38.367 | 10.624 | 6,809 | 35.83 | 10.671 | 64,205 |
No. of children | 1.459 | 1.565 | 6,809 | 1.768 | 1.629 | 64,205 |
Completed primary | 0.873 | 0.333 | 6,801 | 0.52 | 0.500 | 64,006 |
Years of schooling | 9.936 | 4.723 | 6,801 | 4.806 | 4.675 | 64,006 |
Currently employed | 0.26 | 0.439 | 6,809 | 0.488 | 0.500 | 64,205 |
Hindu | 0.718 | 0.450 | 6,809 | 0.822 | 0.382 | 64,205 |
SC or ST | 0.121 | 0.326 | 6,780 | 0.279 | 0.449 | 64,142 |
Rural | 0.277 | 0.447 | 6,809 | 0.645 | 0.479 | 64,205 |
Owns refrigerator | 0.952 | 0.215 | 6,807 | 0.224 | 0.417 | 64,072 |
Owns color television | 0.970 | 0.171 | 6,807 | 0.533 | 0.499 | 64,107 |
Wealth index | 2.602 | 0.288 | 6,809 | 1.791 | 0.605 | 64,205 |
Age, years | 37.514 | 10.631 | 21,650 | 35.409 | 10.651 | 52,105 |
No. of children | 1.524 | 1.533 | 21,650 | 1.838 | 1.67 | 52,105 |
Completed primary | 0.800 | 0.400 | 21,614 | 0.457 | 0.498 | 51,929 |
Years of schooling | 8.375 | 4.844 | 21,614 | 4.066 | 4.342 | 51,929 |
Currently employed | 0.298 | 0.457 | 21,650 | 0.534 | 0.499 | 52,105 |
Hindu | 0.770 | 0.421 | 21,650 | 0.832 | 0.374 | 52,105 |
SC or ST | 0.164 | 0.370 | 21,604 | 0.301 | 0.459 | 52,059 |
Rural | 0.384 | 0.486 | 21,650 | 0.700 | 0.458 | 52,105 |
Owns washing machine | 0.311 | 0.463 | 20,855 | 0.007 | 0.081 | 50,024 |
Owns color television | 0.927 | 0.260 | 21,637 | 0.423 | 0.494 | 52,080 |
Wealth index | 2.449 | 0.352 | 21,650 | 1.638 | 0.558 | 52,105 |
IHDS, India Human Development Survey; SC/ST, scheduled caste/scheduled tribe.
Our variable of interest in the sample of married women is employment status. Since the IHDS does not provide data on LFP, we construct the variable using information on different work categories. The work categories include the following: work on family farm, work tending to the household's animals, family's nonfarm business work, agricultural wage labor, nonagricultural wage labor, and salaried position. Participation in each category is given a value of one if the individual spent ≥240 hours in that activity in the past year and zero otherwise. Employment status is indicated using a dummy variable, equaling one if the woman participates in any of the work categories listed above and zero otherwise. It appears that married women who live in households that own time-saving appliances are
In addition to studying the effect of time-saving appliance ownership on the overall employment rate, we analyze the effect of time-saving appliance ownership on different types of employment, as some employment types are viewed as more valuable than others. For example, Desai et al. (2010) suggest that salaried work (typically found outside the home) is at the top of the job ladder, while manual labor positions (typically found inside the home) are valued much lower. We use IHDS data on each individual's work category, listed above, to classify whether an individual's employment is in work inside or outside the household. Using the definitions provided by the IHDS, we construct the variable work inside the household by combining the three work categories of animal care, farm work, and nonfarm family business work. These are mainly manual labor roles and are typically unpaid or low paid. Similarly, we construct the variable work outside the household for household members who work for payment (wage or salary) by combining the three work categories of agricultural wage labor, nonagricultural wage labor, or salaried position worker. These positions are viewed as more valuable, as they typically come with higher pay than for work inside the household. Our categories follow the categories used in previous literature (such as by Desai et al., 2010).
Table A1 in the Appendix presents the trends in employment rate separately for married and single women in the working-age population. The overall employment rate for married women declined from 52% in 2005 to 48% in 2012. At the same time, the overall employment rate for single women declined from 36% in 2005 to 34% in 2012. However, the table indicates that the decline in employment rate observed for married and single women occurs in all work inside the household job types. Married and single women experience an increase in two of the work roles outside the household (nonagricultural labor and salaried position) between 2005 and 2012, suggesting that the breakdown by employment type in the empirical analysis is warranted.
It is beneficial to examine the differences between married and single women beyond their employment decisions as we will use average appliance ownership rate among Given the higher prevalence of refrigerator ownership, we additionally present differences among women across waves partitioned by refrigerator ownership in Table A2 in the Appendix.
To understand how an appliance affects a household as a whole, we additionally perform our empirical analysis, described in the following section, on a child sample consisting of 50,160 children over the 2 years of data. Descriptive statistics for the child sample are presented in Table 2, which follows the same format as Table 1. We analyze children between the ages of 12 years and 18 years, as the enrollment rate begins to decline from 96% to 92% and the employment rate doubles from 3% to 6% at the age of 12 years in the IHDS data. We observe that primary school completion rate is 10 percentage points higher for children living in households that own an appliance than children living in households that do not own an appliance. In addition, these appliance-owning children are more likely to be enrolled in school and less likely to be employed. Children who live in a household that does not own an appliance are more likely to be from a SC or ST background and from a rural area. Similar to married women, most children who live in a household that owns a washing machine also own a refrigerator and color television. Parents’ ages appear to be similar across children who live in households with and without appliances, but both parents possess a higher number of years of schooling for those children who live in a household that owns an appliance. Not surprisingly, the table suggests a higher index of wealth among appliance owners.
Summary statistics by ownership of appliance: children 12–18 years of age
Completed primary | 0.965 | 0.185 | 3,607 | 0.855 | 0.352 | 45,882 |
Years of schooling | 8.660 | 2.414 | 3,607 | 7.216 | 2.939 | 45,882 |
Currently enrolled | 0.950 | 0.219 | 3,607 | 0.746 | 0.435 | 45,890 |
Currently employed | 0.045 | 0.207 | 3,616 | 0.182 | 0.386 | 45,987 |
Female | 0.468 | 0.499 | 3,616 | 0.488 | 0.500 | 45,987 |
Hindu | 0.683 | 0.465 | 3,616 | 0.784 | 0.411 | 45,987 |
SC or ST | 0.154 | 0.361 | 3,605 | 0.296 | 0.456 | 45,964 |
Rural | 0.264 | 0.441 | 3,616 | 0.650 | 0.477 | 45,987 |
Owns refrigerator | 0.943 | 0.232 | 3,613 | 0.187 | 0.390 | 45,898 |
Owns color television | 0.956 | 0.205 | 3,616 | 0.473 | 0.499 | 45,932 |
Wealth index | 2.582 | 0.299 | 3,616 | 1.726 | 0.596 | 45,987 |
Father's age | 44.610 | 6.680 | 2,492 | 43.820 | 8.402 | 29,733 |
Father's years of schooling | 10.514 | 4.064 | 2,490 | 5.916 | 4.610 | 29,680 |
Mother's age | 39.448 | 7.132 | 2,700 | 38.404 | 8.246 | 31,803 |
Mother's years of schooling | 8.641 | 4.897 | 2,700 | 3.710 | 4.270 | 31,752 |
Completed primary | 0.949 | 0.219 | 12,500 | 0.840 | 0.367 | 38,976 |
Years of schooling | 8.436 | 2.490 | 12,500 | 6.997 | 2.961 | 38,976 |
Currently enrolled | 0.898 | 0.302 | 12,500 | 0.718 | 0.450 | 38,985 |
Currently employed | 0.080 | 0.271 | 12,521 | 0.199 | 0.399 | 39,069 |
Female | 0.483 | 0.500 | 12,521 | 0.489 | 0.500 | 39,069 |
Hindu | 0.731 | 0.443 | 12,521 | 0.792 | 0.406 | 39,069 |
SC or ST | 0.192 | 0.394 | 12,505 | 0.311 | 0.463 | 39,051 |
Rural | 0.380 | 0.485 | 12,521 | 0.695 | 0.460 | 39,069 |
Owns washing machine | 0.284 | 0.451 | 11,983 | 0.006 | 0.074 | 37,528 |
Owns color television | 0.912 | 0.284 | 12,514 | 0.376 | 0.484 | 39,051 |
Wealth index | 2.421 | 0.356 | 12,521 | 1.598 | 0.548 | 39,069 |
Father's age | 44.153 | 7.466 | 8,499 | 43.676 | 8.638 | 24,789 |
Father's years of schooling | 9.305 | 4.282 | 8,492 | 5.277 | 4.416 | 24,737 |
Mother's age | 38.868 | 7.594 | 9,187 | 38.254 | 8.435 | 26,385 |
Mother's years of schooling | 7.058 | 4.856 | 9,181 | 3.118 | 3.911 | 26,337 |
IHDS, India Human Development Survey; SC/ST, scheduled caste/scheduled tribe.
One concern with estimating the effect of household appliances on individual outcomes is that the appliances considered in this analysis are a large purchase and are typically available to wealthier households. These appliances might pick up a wealth effect, whereby wealthier households simultaneously purchase appliances and alter their daily activities. To address this concern, we control for wealth using a wealth index, which uses detailed data and places households in the analysis on a continuous scale of relative wealth. This control assists with distinguishing between the wealth effect and the true effect of the time-saving appliance on LFP and schooling. Wealth indices are commonly used in research in developing countries. The Demographic and Health Surveys (DHS) regularly create wealth indices when studying health services, as “the distribution of health services to the poor can be determined by a wealth index as well as or better than an income or expenditure index. This is because of the lower volatility of wealth as compared with that of income and expenditures” (
We use information on each household's ownership of time-using appliances and household infrastructure to construct a wealth index. A detailed description of the construction of this index is available in Kerr (2019). Briefly, we use a factor analysis approach and assign factor scores to household assets and infrastructure (see Kennedy [2008, p. 200] for an explanation on factor analysis). We then combine the scores by household to rank each household's overall wealth. The time-using assets in our wealth index include the following: air cooler, black- and-white television, bicycle, cable, car, cellphone, chair, computer, cot, generator, laptop, mixer, motorcycle, sewing machine, shoes, telephone, and two sets of clothes. Air conditioner and color television are not included in our wealth index, as these two appliances are included in our instrument, which is described in the following section.
Household infrastructure variables include discrete indicators for access to drinking water and toilet facilities; principal sources for water; and the types of chulha, A chulha is a traditional Indian stove.
We use these infrastructure variables and asset ownership indicator variables to estimate the factor scores for each household in each year of the survey. Again, following Kerr (2019), we sum the factor scores within each household each year to create an index between the values 0 and 3.18, with 0 representing the least wealthy households and 3.18 representing the wealthiest households. We compare the index with available household income to verify that the households that report higher incomes also show higher wealth indices.
Summary statistics presented above suggest differences in married women's employment and children's schooling and employment, which appear to be identified on the basis of ownership of durable appliances. The purpose of the empirical analysis is to ascertain the extent to which this difference is sustained in the multivariate context of Eqs (1) and (3), described in the following section.
We use detailed household data from the IDHS, described in the previous section, to estimate the model and examine the hypothesis that as households purchase time-saving durable goods, married women and children allocate their time away from household work to employment in economic activities or schooling. Employment is captured using a dummy variable, which equals one if the individual is engaged in any of the work categories described in the previous section and zero otherwise. Schooling is captured using a dummy variable, which equals one if the individual is currently enrolled in school and zero otherwise.
An expression that represents the relationship between appliance ownership and the two outcome variables explored in this analysis is given as follows:
Elsewhere in Eq. (1),
Appended to Eq. (1) are latent fixed effects for the district of residence, π
A concern when estimating Eq. (1) is the possibility that appliance ownership is correlated with unobserved factors that determine time allocation. Coen-Pirani et al. (2010) recognize this problem and explain potential biases that exist when using ordinary least squares (OLS) to test the effect of appliance ownership on married women's LFP. Kerr (2019) presents an adapted version of their reasoning, which is relevant to child outcomes. Both papers apply to our analysis, suggesting potential biases such as the following: (1) households with married women in the labor force or children in school are more likely to purchase appliances due to greater utility from their services, or (2) households with strong tastes for home-produced goods might invest heavily in both household work and household appliances. These household preferences are unobserved and are also possible determinants of married women's LFP and children's human capital investment. Stated differently, in the context of child outcomes, households may choose to invest in both schooling and appliances if they have a strong preference for child outcomes and believe that time-saving appliance ownership will relieve children of household duties. Alternatively, negative selection bias may be present if households purchase appliances for reasons other than to reduce household work.
Controlling for variables such as parental and village characteristics will mitigate these challenges as these variables represent measures of parental preferences and availability within the village, but there remains the potential for other unobservables to bias the results obtained from OLS. To address this issue, we use 2SLS estimation to address potential sources of endogeneity. As part of this approach, we use an IV strategy to identify the causal effect of owning a time-saving appliance on both married woman outcomes and child outcomes. For the instrument to be valid, it must affect the potentially endogenous variable (living in a household that owns a time-saving appliance) but have no direct effect on the dependent variable (employment or school enrollment). To determine the effect of household appliance ownership on married women's LFP, Coen-Pirani et al. (2010) instrument a married woman's ownership of an appliance by average ownership rate for that appliance among
Moving to child outcomes, Kerr (2019) adapts Coen-Pirani et al.'s (2010) specification by instrumenting a child's household ownership of a time-saving appliance by the average ownership rate for that same appliance among households
In addition to the PSU-level instruments described above, we follow Kerr (2019) and include two family-specific time-using household assets in our first-stage estimation: family ownership of an air conditioner and color television. The logic underlying these choices is that these two appliances are (1) likely to be associated, on a household basis, with ownership of time-saving appliances such as washing machines and refrigerators and (2) unlikely to exert direct effects on married women's and children's employment or school enrollment. These two appliances are perhaps revealing about other characteristics of the household that might explain their acquisition of time-saving appliances. Results excluding these two family-specific time-using household assets from our instrument are available upon request.
In the instrumented version of the model, we define household time-saving appliance ownership as a function of average PSU ownership and household ownership of two time-using appliances:
The results of the estimation are presented in Tables 3–10. The principal focus of estimation is the parameter β1 in Eqs (1) and (3), the effect of household appliance ownership on married woman and child outcomes. We present the estimates obtained by both OLS and the 2SLS procedure described in Section 4. Recall that OLS results are not believed to be causal. However, they are a useful point of departure and are used to assess our 2SLS results. We partition our child sample by gender to further examine the relationship between appliance ownership and child outcomes. To assess the robustness of our estimates, we present one set of estimates for Eqs (1) and (3) for which appliance ownership is identified by the presence of a washing machine and a second set based on ownership of a refrigerator. We additionally perform a second analysis excluding the two household-specific instruments from our first-stage estimation. However, these results are not included in the main text, as this approach is less conservative than the approach described in Section 4. These results are available on request.
OLS: effect of living in a household that owns an
Owns washing machine | 0.030*** (0.009) | 0.004 (0.018) | ||
Owns refrigerator | −0.001 (0.007) | −0.010 (0.011) | ||
Age | 0.054*** (0.002) | 0.064*** (0.002) | 0.054*** (0.002) | 0.063*** (0.002) |
Age-squared | −0.001*** (0.000) | −0.001*** (0.000) | −0.001*** (0.000) | −0.001*** (0.000) |
Years of schooling | −0.007*** (0.001) | −0.012*** (0.001) | −0.007*** (0.001) | −0.012*** (0.001) |
Rural | 0.244*** (0.014) | 0.043* (0.025) | 0.246*** (0.014) | 0.043* (0.025) |
Hindu | 0.060*** (0.009) | 0.078*** (0.014) | 0.060*** (0.009) | 0.079*** (0.014) |
SC or ST | 0.015** (0.007) | 0.013 (0.009) | 0.014** (0.007) | 0.013 (0.009) |
Wealth index | −0.146*** (0.006) | −0.112*** (0.008) | −0.144*** (0.006) | −0.110*** (0.008) |
No. of children (household) | 0.000 (0.002) | −0.004* (0.002) | −0.000 (0.002) | −0.005** (0.002) |
Distance to nearest town (km) | −0.000 (0.000) | −0.000 (0.000) | ||
Distance to road access (km) | 0.003** (0.001) | 0.003* (0.001) | ||
Phone access | −0.014 (0.011) | −0.013 (0.011) | ||
Distance to railway station (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Distance to market (km) | 0.002** (0.001) | 0.002*** (0.001) | ||
Distance to secondary school (km) | 0.002*** (0.001) | 0.003*** (0.001) | ||
Distance to higher secondary school (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Average |
0.096 | 0.046 | 0.294 | 0.191 |
Average employment | 0.466 | 0.619 | 0.465 | 0.619 |
N | 70,263 | 38,109 | 73,078 | 39,398 |
0.303 | 0.228 | 0.302 | 0.225 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
OLS, ordinary least squares; SC/ST, scheduled caste/scheduled tribe.
As a point of departure, Table 3 presents the OLS estimates of Eq. (1), with appliance ownership defined as the presence of a washing machine in Columns 1 and 2 and the presence of a refrigerator in Columns 3 and 4. Living in a household that owns a washing machine increases married women's probability of employment by 3 percentage points (Column 1 of Table 3). Refrigerator ownership is found to be insignificant in married women's employment decisions (Column 3 of Table 3). Estimates obtained using PSU-level controls (which estimate the effect of appliance ownership on married women living in villages only) indicate insignificant changes in employment with ownership of either appliance (Columns 2 and 4 of Table 3).
Some of the control variables provide further insight into the Indian setting, where women appear to be reducing their LFP when a concomitant increase in other wealth-indicating characteristics occurs, which is inconsistent with general intuition. Our results suggest that women The directions of these effects are plausible and consistent with previous literature on female labor supply in India. See Bhargava (2020), Klasen and Pieters (2015), Afridi et al. (2018), and Neff et al. (2012), among others.
Tables 4 and 5 present the OLS estimates of Eq. (1) for the child sample for washing machines and refrigerators, respectively. Columns 1–6 and 7–12 indicate the results for enrollment and employment, respectively. We find an insignificant increase in the probability of being enrolled and an insignificant decrease in the probability of being employed in nearly all specifications for both appliances. The wealth index is highly significant in all models and for both genders. Unlike the married-female sample, the direction of the effect follows intuition and is plausible: households characterized by greater wealth tend to (1) more actively enroll children in school and (2) engage children to a lesser extent in market work. Among the controls, parents’ years of schooling are significant in all cases, with signs that largely match those of the wealth index. The remaining controls are only sporadically significant, yet their inclusion is useful as a general principle in isolating the key independent variable, appliance ownership, after controlling for household and PSU characteristics.
OLS: effect of living in a household that owns a washing machine on children's outcomes
Owns washing machine | 0.001 (0.008) | 0.019 (0.013) | −0.004 (0.011) | 0.018 (0.015) | 0.007 (0.010) | 0.024 (0.019) | −0.004 (0.007) | −0.025* (0.014) | −0.006 (0.010) | −0.024 (0.020) | −0.004 (0.009) | −0.023 (0.018) |
Age | 0.142*** (0.017) | 0.168*** (0.024) | 0.167*** (0.022) | 0.202*** (0.030) | 0.108*** (0.027) | 0.129*** (0.038) | −0.064*** (0.017) | −0.062** (0.025) | −0.096*** (0.023) | −0.092*** (0.032) | 0.012 (0.023) | 0.016 (0.035) |
Age–squared | −0.007*** (0.001) | −0.008*** (0.001) | −0.008*** (0.001) | −0.009*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | 0.004*** (0.001) | 0.004*** (0.001) | 0.005*** (0.001) | 0.005*** (0.001) | 0.000 (0.001) | 0.001 (0.001) |
Female | −0.016*** (0.005) | −0.032*** (0.008) | −0.055*** (0.005) | −0.051*** (0.007) | ||||||||
Rural | 0.030*** (0.008) | −0.034* (0.018) | 0.046*** (0.011) | −0.019 (0.023) | 0.006 (0.010) | −0.048 (0.031) | 0.058*** (0.009) | 0.009 (0.023) | 0.054*** (0.011) | 0.017 (0.031) | 0.062*** (0.010) | 0.003 (0.030) |
Hindu | 0.080*** (0.010) | 0.073*** (0.014) | 0.077*** (0.011) | 0.065*** (0.014) | 0.085*** (0.012) | 0.084*** (0.020) | −0.007 (0.006) | −0.005 (0.010) | −0.019** (0.009) | −0.020 (0.015) | 0.003 (0.007) | 0.010 (0.013) |
SC or ST | −0.011* (0.006) | −0.020*** (0.008) | −0.018** (0.008) | −0.027*** (0.009) | −0.003 (0.008) | −0.010 (0.011) | −0.009 (0.006) | −0.007 (0.008) | −0.007 (0.007) | −0.008 (0.010) | −0.013* (0.007) | −0.009 (0.011) |
Wealth index | 0.096*** (0.008) | 0.079*** (0.011) | 0.096*** (0.009) | 0.076*** (0.011) | 0.095*** (0.010) | 0.083*** (0.014) | −0.067*** (0.006) | −0.054*** (0.008) | −0.065*** (0.008) | −0.050*** (0.011) | −0.068*** (0.008) | −0.058*** (0.011) |
Father's age | −0.003*** (0.000) | −0.002*** (0.001) | −0.002*** (0.001) | −0.002* (0.001) | −0.004*** (0.001) | −0.003*** (0.001) | 0.001** (0.000) | 0.001 (0.001) | 0.001** (0.001) | 0.000 (0.001) | 0.001* (0.001) | 0.001 (0.001) |
Mother's age | 0.000 (0.000) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | −0.000 (0.000) | 0.000 (0.001) | −0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | −0.000 (0.001) |
Father's years of schooling | 0.012*** (0.001) | 0.012*** (0.001) | 0.012*** (0.001) | 0.012*** (0.001) | 0.013*** (0.001) | 0.013*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.005*** (0.001) |
Mother's years of schooling | 0.006*** (0.001) | 0.006*** (0.001) | 0.007*** (0.001) | 0.006*** (0.001) | 0.006*** (0.001) | 0.005*** (0.001) | −0.004*** (0.001) | −0.005*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.004*** (0.001) |
Distance to nearest town (km) | 0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | −0.001 (0.000) | −0.000 (0.001) | −0.001** (0.001) | ||||||
Distance to road access (km) | 0.001 (0.002) | 0.003 (0.002) | −0.000 (0.002) | 0.000 (0.002) | −0.002 (0.003) | 0.002 (0.002) | ||||||
Phone access | 0.003 (0.011) | 0.001 (0.013) | 0.010 (0.015) | −0.030** (0.012) | −0.027* (0.014) | −0.029* (0.016) | ||||||
Distance to railway station (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | 0.000 (0.000) | −0.000 (0.000) | ||||||
Distance to market (km) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.000 (0.001) | 0.002* (0.001) | ||||||
Distance to secondary school (km) | −0.001 (0.001) | −0.000 (0.001) | −0.002 (0.001) | 0.002** (0.001) | 0.002 (0.001) | 0.002* (0.001) | ||||||
Distance to higher secondary school (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.001) | 0.000 (0.000) | ||||||
Average washing machine ownership | 0.079 | 0.037 | 0.076 | 0.038 | 0.083 | 0.037 | 0.079 | 0.037 | 0.076 | 0.038 | 0.083 | 0.037 |
Average |
0.819 | 0.806 | 0.813 | 0.806 | 0.827 | 0.807 | 0.152 | 0.194 | 0.184 | 0.226 | 0.110 | 0.152 |
N | 30,348 | 17,127 | 17,168 | 9,744 | 13,176 | 7,378 | 30,363 | 17,132 | 17,177 | 9,748 | 13,182 | 7,379 |
0.248 | 0.245 | 0.249 | 0.246 | 0.281 | 0.290 | 0.170 | 0.167 | 0.196 | 0.196 | 0.163 | 0.166 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
OLS, ordinary least squares; SC/ST, scheduled caste/scheduled tribe.
The estimates presented in Tables 4 and 5 suggest that time-saving household appliances do not consistently exert significant effects on employment or enrollment for children. However, the important caveat is the possible endogeneity of appliance ownership. We address this by using a 2SLS estimation of Eq. (1), the results of which are discussed in the following sections.
OLS: effect of living in a household that owns a refrigerator on children's outcomes
Owns refrigerator | 0.017** (0.007) | 0.010 (0.010) | 0.015* (0.009) | 0.003 (0.011) | 0.018** (0.009) | 0.018 (0.014) | 0.005 (0.006) | −0.001 (0.010) | 0.006 (0.007) | 0.001 (0.013) | 0.002 (0.008) | −0.005 (0.015) |
Age | 0.149*** (0.017) | 0.169*** (0.024) | 0.173*** (0.022) | 0.201*** (0.031) | 0.113*** (0.027) | 0.133*** (0.037) | −0.067*** (0.017) | −0.068*** (0.024) | −0.100*** (0.022) | −0.099*** (0.032) | 0.011 (0.022) | 0.015 (0.035) |
Age–squared | −0.007*** (0.001) | −0.008*** (0.001) | −0.008*** (0.001) | −0.009*** (0.001) | −0.006*** (0.001) | −0.007*** (0.001) | 0.004*** (0.001) | 0.004*** (0.001) | 0.005*** (0.001) | 0.006*** (0.001) | 0.000 (0.001) | 0.001 (0.001) |
Female | −0.015*** (0.005) | −0.030*** (0.008) | −0.056*** (0.005) | −0.051*** (0.007) | ||||||||
Rural | 0.029*** (0.008) | −0.036* (0.018) | 0.044*** (0.011) | −0.022 (0.023) | 0.005 (0.010) | −0.047 (0.031) | 0.059*** (0.008) | 0.010 (0.023) | 0.056*** (0.010) | 0.016 (0.031) | 0.061*** (0.010) | 0.007 (0.029) |
Hindu | 0.080*** (0.009) | 0.074*** (0.014) | 0.077*** (0.011) | 0.066*** (0.014) | 0.085*** (0.012) | 0.084*** (0.020) | −0.004 (0.006) | −0.001 (0.011) | −0.014 (0.009) | −0.013 (0.015) | 0.003 (0.007) | 0.010 (0.013) |
SC or ST | −0.010* (0.006) | −0.019** (0.008) | −0.016** (0.008) | −0.025** (0.010) | −0.003 (0.008) | −0.010 (0.011) | −0.008 (0.005) | −0.007 (0.008) | −0.007 (0.007) | −0.007 (0.010) | −0.011 (0.007) | −0.007 (0.011) |
Wealth index | 0.090*** (0.008) | 0.079*** (0.011) | 0.092*** (0.009) | 0.080*** (0.012) | 0.088*** (0.011) | 0.078*** (0.015) | −0.071*** (0.006) | −0.059*** (0.008) | −0.071*** (0.008) | −0.058*** (0.011) | −0.069*** (0.008) | −0.058*** (0.011) |
Father's age | −0.003*** (0.000) | −0.002*** (0.001) | −0.002*** (0.001) | −0.002* (0.001) | −0.004*** (0.001) | −0.003*** (0.001) | 0.001*** (0.000) | 0.001 (0.001) | 0.001** (0.001) | 0.000 (0.001) | 0.001** (0.001) | 0.001 (0.001) |
Mother's age | 0.000 (0.000) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | −0.000 (0.000) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | −0.000 (0.001) |
Father's years of schooling | 0.012*** (0.001) | 0.012*** (0.001) | 0.011*** (0.001) | 0.012*** (0.001) | 0.012*** (0.001) | 0.013*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.005*** (0.001) |
Mother's years of schooling | 0.006*** (0.001) | 0.006*** (0.001) | 0.007*** (0.001) | 0.007*** (0.001) | 0.006*** (0.001) | 0.005*** (0.001) | −0.004*** (0.001) | −0.005*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.004*** (0.001) |
Distance to nearest town (km) | 0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | −0.001 (0.000) | −0.000 (0.001) | −0.001* (0.001) | ||||||
Distance to road access (km) | 0.001 (0.001) | 0.002 (0.002) | −0.000 (0.002) | 0.000 (0.002) | −0.002 (0.003) | 0.002 (0.002) | ||||||
Phone access | 0.003 (0.011) | 0.000 (0.012) | 0.009 (0.015) | −0.032*** (0.012) | −0.029** (0.014) | −0.030* (0.016) | ||||||
Distance to railway station (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | 0.000 (0.000) | −0.000 (0.000) | ||||||
Distance to market (km) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.000 (0.001) | 0.002 (0.001) | ||||||
Distance to secondary school (km) | −0.001 (0.001) | −0.000 (0.001) | −0.002 (0.001) | 0.002** (0.001) | 0.002* (0.001) | 0.003** (0.001) | ||||||
Distance to higher secondary school (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.001) | 0.000 (0.000) | ||||||
Average refrigerator ownership | 0.261 | 0.166 | 0.250 | 0.164 | 0.276 | 0.169 | 0.261 | 0.166 | 0.250 | 0.163 | 0.276 | 0.169 |
Average |
0.819 | 0.807 | 0.812 | 0.805 | 0.828 | 0.808 | 0.151 | 0.194 | 0.183 | 0.227 | 0.108 | 0.151 |
N | 31,376 | 17,518 | 17,833 | 10,009 | 13,539 | 7,506 | 31,391 | 17,523 | 17,842 | 10,013 | 13,545 | 7,507 |
0.247 | 0.244 | 0.248 | 0.246 | 0.280 | 0.289 | 0.170 | 0.167 | 0.195 | 0.195 | 0.162 | 0.165 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
OLS, ordinary least squares; SC/ST, scheduled caste/scheduled tribe.
To address the concern of appliance ownership being endogenous, we begin our 2SLS estimation by estimating Eq. (2), which regresses household ownership of time-saving appliances for married women against average PSU ownership of time-saving appliances for single women, ownership of an air conditioner, ownership of a color television, and the set of controls included in our OLS estimation. The model is estimated separately for ownership of washing machines and refrigerators.
Estimates of Eq. (2) are compressed to the first four rows of Table 6 and suggest that the likelihood of owning either appliance increases for married women when the average PSU ownership of the same appliance increases for single women living in the same PSU. Tables including first-stage results that present all controls are available upon request.
2SLS: effect of living in a household that owns an
Owns washing machine (single women) | 0.501*** (0.017) | 0.485*** (0.031) | ||
Owns refrigerator (single women) | 0.376*** (0.013) | 0.352*** (0.020) | ||
Owns air conditioner | 0.437*** (0.024) | 0.523*** (0.0450) | 0.158*** (0.019) | 0.223*** (0.031) |
Owns color television | 0.003 (0.004) | 0.000 (0.004) | 0.121*** (0.008) | 0.100*** (0.011) |
638.36 | 146.45 | 495.87 | 193.80 | |
Owns washing machine | 0.030 (0.035) | −0.074 (0.061) | ||
Owns refrigerator | −0.102*** (0.031) | −0.097* (0.055) | ||
Age | 0.054*** (0.002) | 0.064*** (0.002) | 0.054*** (0.002) | 0.064*** (0.002) |
Age–squared | −0.001*** (0.000) | −0.001*** (0.000) | −0.001*** (0.000) | −0.001*** (0.000) |
Years of schooling | −0.007*** (0.001) | −0.012*** (0.001) | −0.005*** (0.001) | −0.011*** (0.001) |
Rural | 0.242*** (0.014) | 0.041 (0.025) | 0.239*** (0.014) | 0.036 (0.026) |
Hindu | 0.060*** (0.009) | 0.075*** (0.014) | 0.057*** (0.009) | 0.074*** (0.014) |
SC or ST | 0.015** (0.007) | 0.012 (0.009) | 0.013* (0.007) | 0.010 (0.009) |
Wealth index | −0.140*** (0.007) | −0.102*** (0.009) | −0.107*** (0.010) | −0.082*** (0.016) |
No. of children (household) | 0.000 (0.002) | −0.004* (0.002) | 0.001 (0.002) | −0.004* (0.002) |
Distance to nearest town (km) | −0.000 (0.000) | −0.000 (0.000) | ||
Distance to road access (km) | 0.003** (0.001) | 0.003* (0.001) | ||
Phone access | −0.014 (0.011) | −0.016 (0.011) | ||
Distance to railway station (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Distance to market (km) | 0.002** (0.001) | 0.001** (0.001) | ||
Distance to secondary school (km) | 0.002*** (0.001) | 0.002*** (0.001) | ||
Distance to higher secondary school (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Average |
0.096 | 0.046 | 0.294 | 0.194 |
Average employment | 0.466 | 0.619 | 0.467 | 0.619 |
N | 70,160 | 38,059 | 70,221 | 38,071 |
0.303 | 0.227 | 0.298 | 0.225 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
2SLS, two–stage least squares; SC/ST, scheduled caste/scheduled tribe.
The remaining rows of Table 6 present our 2SLS results using Eq. (3) for married women. Columns 1 and 2 suggest insignificant effects of washing machine ownership on married women's employment. In contrast, we find that married women living in households that own a refrigerator decrease their probability of employment by 10 percentage points (Columns 3 and 4 of Table 6). This finding is significant in both the full sample (Column 3) and the village sample (Column 4) and does not follow most intuition or previous studies. The results suggest that time-saving appliances allow women to For robustness, we generate another index by regressing appliance ownership on all household assets/infrastructure variables from the original wealth index, and then, we use the residual in our final specification as an alternative index to control for household wealth. The coefficients on washing machine and refrigerator ownership estimated using this strategy are larger than our main results and are highly significant. These results further validate our results and are available upon request. We thank an anonymous reviewer for suggesting this robustness check.
Before moving on to the child sample, we split our sample of married women on the basis of their education to further distinguish the human capital effect. Table A3 in the Appendix presents these results, following the same empirical strategy as above, separately for women who did not complete primary education and for women with an education level of primary and above. The results presented in Table A3 in the Appendix are consistent with those presented in Table 6, suggesting insignificant effects of washing machine ownership for both education levels and significant effects of refrigerator ownership on the probability of married women's employment for both education levels. The coefficient is larger for the sample of married women who did not complete primary (−18.4% in Column 5 compared to −13.1% in Column 7 of Table A3 in the Appendix), suggesting that these time-saving appliances have a larger impact on the probability of being employed for less-educated married women. We thank an anonymous reviewer for suggesting this robustness check.
First-stage estimates using Eq. (2) for the child sample are presented in the first three rows of Tables 7 and 8. Similar to the married women sample, the results suggest that ownership of time-saving appliances among households with children increases as the average PSU ownership among households with no children in the same PSU increases.
2SLS: effect of living in a household that owns a washing machine on children's outcomes
Owns washing machine (no children) | 0.379*** (0.027) | 0.406*** (0.042) | 0.405*** (0.026) | 0.454*** (0.045) | 0.348*** (0.034) | 0.348*** (0.055) | 0.379*** (0.027) | 0.406*** (0.042) | 0.405*** (0.026) | 0.454*** (0.045) | 0.348*** (0.034) | 0.348*** (0.055) |
Owns air conditioner | 0.483*** (0.028) | 0.496*** (0.075) | 0.460*** (0.035) | 0.499*** (0.093) | 0.515*** (0.033) | 0.504*** (0.091) | 0.484*** (0.028) | 0.496*** (0.075) | 0.460*** (0.035) | 0.499*** (0.093) | 0.515*** (0.033) | 0.504*** (0.091) |
Owns color television | −0.002 (0.005) | 0.004 (0.005) | 0.005 (0.005) | 0.008* (0.005) | −0.009 (0.007) | −0.002 (0.006) | −0.002 (0.005) | 0.004 (0.005) | 0.005 (0.005) | 0.008* (0.005) | −0.009 (0.007) | −0.001 (0.006) |
108.67 | 17.72 | 156.08 | 50.09 | 175.99 | 30.48 | 207.77 | 55.00 | 156.10 | 50.08 | 176.29 | 30.48 | |
Owns washing machine | −0.014 (0.022) | 0.034 (0.062) | −0.016 (0.028) | 0.022 (0.064) | −0.020 (0.030) | 0.036 (0.087) | 0.009 (0.021) | −0.055 (0.067) | 0.004 (0.032) | −0.002 (0.074) | −0.003 (0.022) | −0.162* (0.087) |
Age | 0.143*** (0.017) | 0.169*** (0.024) | 0.168*** (0.022) | 0.203*** (0.030) | 0.109*** (0.027) | 0.129*** (0.038) | −0.065*** (0.017) | −0.063** (0.025) | −0.098*** (0.023) | −0.094*** (0.032) | 0.012 (0.023) | 0.019 (0.035) |
Age–squared | −0.007*** (0.001) | −0.008*** (0.001) | −0.008*** (0.001) | −0.009*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | 0.004*** (0.001) | 0.004*** (0.001) | 0.005*** (0.001) | 0.005*** (0.001) | 0.000 (0.001) | 0.001 (0.001) |
Female | −0.016*** (0.005) | −0.032*** (0.008) | −0.055*** (0.005) | −0.051*** (0.007) | ||||||||
Rural | 0.029*** (0.008) | −0.033* (0.019) | 0.045*** (0.012) | −0.019 (0.023) | 0.004 (0.011) | −0.047 (0.033) | 0.058*** (0.009) | 0.006 (0.024) | 0.054*** (0.010) | 0.017 (0.031) | 0.062*** (0.010) | −0.009 (0.034) |
Hindu | 0.081*** (0.010) | 0.074*** (0.014) | 0.078*** (0.011) | 0.066*** (0.015) | 0.085*** (0.013) | 0.086*** (0.020) | −0.007 (0.006) | −0.007 (0.011) | −0.019** (0.009) | −0.021 (0.015) | 0.003 (0.007) | 0.006 (0.014) |
SC or ST | −0.011* (0.006) | −0.020*** (0.008) | −0.018** (0.008) | −0.027*** (0.009) | −0.002 (0.008) | −0.010 (0.012) | −0.009 (0.006) | −0.008 (0.008) | −0.007 (0.007) | −0.008 (0.010) | −0.013* (0.007) | −0.010 (0.011) |
Wealth index | 0.090*** (0.007) | 0.072*** (0.010) | 0.090*** (0.008) | 0.069*** (0.011) | 0.091*** (0.010) | 0.076*** (0.015) | −0.065*** (0.006) | −0.050*** (0.009) | −0.063*** (0.008) | −0.050*** (0.011) | −0.064*** (0.008) | −0.046*** (0.012) |
Father's age | −0.003*** (0.000) | −0.002*** (0.001) | −0.002*** (0.001) | −0.002* (0.001) | −0.004*** (0.001) | −0.003*** (0.001) | 0.001** (0.000) | 0.001 (0.001) | 0.001* (0.001) | 0.000 (0.001) | 0.001* (0.001) | 0.001 (0.001) |
Mother's age | 0.000 (0.000) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | −0.000 (0.000) | 0.000 (0.001) | −0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | 0.000 (0.001) |
Father's years of schooling | 0.012*** (0.001) | 0.013*** (0.001) | 0.012*** (0.001) | 0.012*** (0.001) | 0.013*** (0.001) | 0.013*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.005*** (0.001) |
Mother's years of schooling | 0.007*** (0.001) | 0.006*** (0.001) | 0.007*** (0.001) | 0.007*** (0.001) | 0.006*** (0.001) | 0.005*** (0.001) | −0.004*** (0.001) | −0.005*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.003*** (0.001) | −0.003** (0.001) |
Distance to nearest town (km) | 0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | −0.001 (0.000) | −0.000 (0.001) | −0.001** (0.001) | ||||||
Distance to road access (km) | 0.002 (0.002) | 0.003 (0.002) | −0.000 (0.002) | 0.000 (0.002) | −0.002 (0.003) | 0.002 (0.002) | ||||||
Phone access | 0.003 (0.011) | 0.001 (0.013) | 0.011 (0.015) | −0.030** (0.012) | −0.028** (0.014) | −0.029* (0.016) | ||||||
Distance to railway station (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | 0.000 (0.000) | −0.000 (0.000) | ||||||
Distance to market (km) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.000 (0.001) | 0.002* (0.001) | ||||||
Distance to secondary school (km) | −0.000 (0.001) | −0.000 (0.001) | −0.002 (0.001) | 0.002** (0.001) | 0.002 (0.001) | 0.002* (0.001) | ||||||
Distance to higher secondary school (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.001) | 0.000 (0.000) | ||||||
Average washing machine ownership | 0.079 | 0.037 | 0.076 | 0.038 | 0.083 | 0.037 | 0.079 | 0.037 | 0.076 | 0.038 | 0.083 | 0.037 |
Average |
0.819 | 0.806 | 0.812 | 0.806 | 0.827 | 0.807 | 0.152 | 0.194 | 0.185 | 0.226 | 0.110 | 0.152 |
N | 30,322 | 17,119 | 17,150 | 9,737 | 13,168 | 7,377 | 30,337 | 17,124 | 17,159 | 9,741 | 13,174 | 7,378 |
0.357 | 0.245 | 0.248 | 0.246 | 0.281 | 0.290 | 0.170 | 0.167 | 0.196 | 0.196 | 0.163 | 0.162 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
2SLS, two–stage least squares; SC/ST, scheduled caste/scheduled tribe.
The remaining rows of Tables 7 and 8 present the estimates of Eq. (3) for ownership of a washing machine and a refrigerator, respectively. Table 7 is consistent with the findings presented in Table 6, suggesting that washing machine ownership does not significantly alter enrollment or employment decisions for children in the sample. One notable difference is found in the village sample of female children. These individuals experience a 16.2 percentage point decrease in the probability of being employed when living in a household that owns a washing machine, suggesting that washing machines alleviate some of the employment opportunities or needs for older female children. An analysis using time-use data would be beneficial to further understand why female children experience this significant change. However, time-use data are not available in this data set.
Observing refrigerator ownership in Table 8, we find statistically significant effects on both enrollment and employment decisions for children. Beginning with enrollment, we observe a 15.3 percentage point increase in the probability of being enrolled when a child lives in a household that owns a refrigerator (Column 1 of Table 8). These significant results hold when the sample is partitioned by gender (Columns 3 and 5). We find that the effect of refrigerator ownership on enrollment is larger for males than for females. Males (females) living in a household that owns a refrigerator experience an increase of 17.5 (12.7) percentage points in the probability of being enrolled. These refrigerator-owning children also experience a 9.1 percentage point decrease in the probability of participating in the labor force (Column 7 of Table 8). Again, the results hold when the sample is partitioned by gender (Columns 9 and 11). Females, however, experience a larger decrease in the probability of employment than males.
2SLS: effect of living in a household that owns a refrigerator on children's outcomes
Owns refrigerator (no children) | 0.290*** (0.019) | 0.268*** (0.030) | 0.310*** (0.022) | 0.285*** (0.034) | 0.264*** (0.022) | 0.246*** (0.035) | 0.290*** (0.019) | 0.267*** (0.030) | 0.310*** (0.022) | 0.285*** (0.034) | 0.265*** (0.022) | 0.246*** (0.035) |
Owns air conditioner | 0.167*** (0.028) | 0.217*** (0.052) | 0.159*** (0.033) | 0.192** (0.077) | 0.174*** (0.028) | 0.256*** (0.052) | 0.167*** (0.028) | 0.217*** (0.052) | 0.160*** (0.033) | 0.192** (0.077) | 0.174*** (0.028) | 0.256*** (0.052) |
Owns color television | 0.092*** (0.009) | 0.073*** (0.011) | 0.101*** (0.010) | 0.080*** (0.013) | 0.082*** (0.011) | 0.066*** (0.013) | 0.092*** (0.009) | 0.073*** (0.011) | 0.101*** (0.010) | 0.080*** (0.013) | 0.083*** (0.011) | 0.066*** (0.013) |
154.95 | 58.22 | 165.63 | 49.05 | 83.00 | 35.51 | 156.18 | 58.14 | 166.03 | 49.01 | 84.50 | 35.49 | |
Own refrigerator | 0.153*** (0.034) | 0.150*** (0.056) | 0.175*** (0.037) | 0.161** (0.062) | 0.127** (0.049) | 0.135* (0.076) | −0.091** (0.036) | −0.065 (0.077) | −0.075* (0.042) | −0.014 (0.081) | −0.106** (0.044) | −0.133 (0.098) |
Age | 0.142*** (0.018) | 0.169*** (0.024) | 0.163*** (0.023) | 0.199*** (0.031) | 0.112*** (0.027) | 0.134*** (0.038) | −0.065*** (0.017) | −0.065*** (0.025) | −0.096*** (0.023) | −0.095*** (0.032) | 0.009 (0.023) | 0.011 (0.035) |
Age–squared | −0.007*** (0.001) | −0.008*** (0.001) | −0.007*** (0.001) | −0.009*** (0.001) | −0.006*** (0.001) | −0.007*** (0.001) | 0.004*** (0.001) | 0.004*** (0.001) | 0.005*** (0.001) | 0.005*** (0.001) | 0.001 (0.001) | 0.001 (0.001) |
Female | −0.016*** (0.005) | −0.030*** (0.008) | −0.056*** (0.005) | −0.051*** (0.007) | ||||||||
Rural | 0.032*** (0.009) | −0.024 (0.018) | 0.047*** (0.012) | −0.010 (0.023) | 0.008 (0.011) | −0.037 (0.031) | 0.055*** (0.009) | 0.005 (0.024) | 0.051*** (0.010) | 0.016 (0.032) | 0.059*** (0.010) | −0.009 (0.030) |
Hindu | 0.086*** (0.010) | 0.079*** (0.014) | 0.085*** (0.011) | 0.072*** (0.015) | 0.090*** (0.012) | 0.090*** (0.020) | −0.010 (0.007) | −0.008 (0.011) | −0.021** (0.009) | −0.020 (0.016) | 0.000 (0.008) | 0.004 (0.015) |
SC or ST | −0.010 (0.006) | −0.017** (0.008) | −0.017** (0.008) | −0.024** (0.010) | −0.001 (0.008) | −0.007 (0.012) | −0.010* (0.006) | −0.009 (0.008) | −0.007 (0.007) | −0.009 (0.010) | −0.014* (0.008) | −0.012 (0.011) |
Wealth index | 0.048*** (0.011) | 0.038** (0.016) | 0.042*** (0.012) | 0.032* (0.018) | 0.055*** (0.016) | 0.046* (0.023) | −0.040*** (0.010) | −0.038** (0.019) | −0.043*** (0.013) | −0.046** (0.021) | −0.035*** (0.014) | −0.025 (0.024) |
Father's age | −0.003*** (0.000) | −0.002*** (0.001) | −0.002*** (0.001) | −0.001* (0.001) | −0.004*** (0.001) | −0.003*** (0.001) | 0.001** (0.000) | 0.001 (0.001) | 0.001** (0.001) | 0.000 (0.001) | 0.001* (0.001) | 0.001 (0.001) |
Mother's age | 0.000 (0.000) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.001) | −0.000 (0.001) | 0.000 (0.000) | 0.001 (0.001) | −0.000 (0.001) | 0.001 (0.001) | 0.000 (0.001) | −0.000 (0.001) |
Father's years of schooling | 0.011*** (0.001) | 0.012*** (0.001) | 0.010*** (0.001) | 0.011*** (0.001) | 0.011*** (0.001) | 0.012*** (0.001) | −0.004*** (0.001) | −0.005*** (0.001) | −0.005*** (0.001) | −0.006*** (0.001) | −0.002** (0.001) | −0.004** (0.001) |
Mother's years of schooling | 0.005*** (0.001) | 0.005*** (0.001) | 0.005*** (0.001) | 0.005*** (0.001) | 0.004*** (0.001) | 0.004*** (0.001) | −0.003*** (0.001) | −0.004*** (0.001) | −0.004*** (0.001) | −0.006*** (0.002) | −0.001 (0.001) | −0.003* (0.001) |
Distance to nearest town (km) | 0.000 (0.001) | 0.001 (0.001) | −0.000 (0.001) | −0.001 (0.000) | −0.000 (0.001) | −0.001** (0.001) | ||||||
Distance to road access (km) | 0.001 (0.002) | 0.002 (0.002) | 0.000 (0.002) | 0.000 (0.002) | −0.001 (0.003) | 0.002 (0.002) | ||||||
Phone access | 0.006 (0.011) | 0.003 (0.013) | 0.013 (0.015) | −0.032*** (0.012) | −0.028** (0.014) | −0.031** (0.016) | ||||||
Distance to railway station (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | 0.000 (0.000) | −0.000 (0.000) | ||||||
Distance to market (km) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | 0.001 (0.001) | −0.000 (0.001) | 0.001 (0.001) | ||||||
Distance to secondary school (km) | −0.000 (0.001) | −0.000 (0.001) | −0.001 (0.001) | 0.002** (0.001) | 0.002 (0.001) | 0.002* (0.001) | ||||||
Distance to higher secondary school (km) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.000) | −0.000 (0.001) | 0.000 (0.000) | ||||||
Average refrigerator ownership | 0.260 | 0.168 | 0.249 | 0.166 | 0.274 | 0.171 | 0.260 | 0.168 | 0.249 | 0.166 | 0.274 | 0.171 |
Average |
0.819 | 0.806 | 0.812 | 0.806 | 0.827 | 0.806 | 0.152 | 0.194 | 0.185 | 0.226 | 0.110 | 0.152 |
N | 30,370 | 17,121 | 17,183 | 9,743 | 13,183 | 7,375 | 30,385 | 17,126 | 17,192 | 9,747 | 13,189 | 7,376 |
0.234 | 0.235 | 0.231 | 0.234 | 0.272 | 0.283 | 0.162 | 0.165 | 0.191 | 0.196 | 0.150 | 0.155 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
2SLS, two–stage least squares; SC/ST, scheduled caste/scheduled tribe.
It is interesting to note that, while the full child sample presents significant findings on refrigerator ownership, performing the analysis on villages provides suggestive evidence that purchasing an appliance is not sufficient in altering labor force decisions for children (Columns 8, 10, and 12 of Table 8). One reason for this finding may be that refrigerator ownership is less prevalent in rural villages (17% ownership in villages compared to 26% ownership in the full sample). We do find a statistically significant increase in the effect of refrigerator ownership on enrollment for the village sample (Columns 2, 4, and 6 of Table 8). These results hold when partitioning the sample by gender. Males and females from the village sample experience a 16.1 percentage point and 13.5 percentage point increase, respectively, in the probability of being enrolled if they live in a household that owns a refrigerator.
Overall, results using the child sample indicate that refrigerator ownership significantly alters both education and employment decisions for older children. When focusing on village-level data, refrigerators increase the probability of enrollment but do not initiate a significant change in employment decisions for older children. Washing machines do not cause similar, significant changes in enrollment or employment decisions.
The results from the married women sample seem perplexing at first, as most research finds appliance ownership to be liberating, allowing women to reduce household work and enter the labor force. For example, see Coen-Pirani et al. (2010) and Greenwood et al. (2005). Tables A4 and A5 in the Appendix show the results for the effect of washing machine and refrigerator ownership, respectively, on the detailed employment categories for married women.
2SLS: effect of living in a household that owns a washing machine on married women's work
Owns washing machine (single women) | 0.501*** (0.0171) | 0.485*** (0.0310) | 0.501*** (0.0171) | 0.485*** (0.0310) |
Owns air conditioner | 0.437*** (0.0239) | 0.523*** (0.0450) | 0.437*** (0.0239) | 0.523*** (0.0450) |
Owns color television | 0.00278 (0.00409) | 0.0000 (0.00407) | 0.00278 (0.00409) | 0.0000 (0.00407) |
638.36 | 146.45 | 638.36 | 146.45 | |
Owns washing machine | −0.068** (0.027) | −0.147** (0.067) | 0.154*** (0.023) | 0.080** (0.033) |
Age | 0.037*** (0.002) | 0.050*** (0.002) | 0.035*** (0.001) | 0.039*** (0.002) |
Age–squared | −0.000*** (0.000) | −0.001*** (0.000) | −0.000*** (0.000) | −0.001*** (0.000) |
Years of schooling | −0.008*** (0.001) | −0.010*** (0.001) | −0.003*** (0.001) | −0.007*** (0.001) |
Rural | 0.286*** (0.014) | 0.076*** (0.024) | 0.014 (0.010) | −0.014 (0.019) |
Hindu | 0.054*** (0.009) | 0.088*** (0.016) | 0.027*** (0.006) | 0.013 (0.009) |
SC or ST | −0.064*** (0.008) | −0.078*** (0.010) | 0.113*** (0.008) | 0.137*** (0.009) |
Wealth index | −0.062*** (0.007) | −0.029*** (0.009) | −0.140*** (0.007) | −0.140*** (0.009) |
No. of children (household) | 0.004** (0.002) | −0.000 (0.002) | −0.005*** (0.001) | −0.005*** (0.002) |
Distance to nearest town (km) | −0.001 (0.000) | 0.000 (0.000) | ||
Distance to road access (km) | 0.003** (0.002) | 0.002 (0.001) | ||
Phone access | −0.043*** (0.013) | 0.016 (0.010) | ||
Distance to railway station (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Distance to market (km) | 0.003*** (0.001) | −0.002*** (0.001) | ||
Distance to secondary school (km) | 0.003*** (0.001) | 0.000 (0.001) | ||
Distance to higher secondary school (km) | 0.000 (0.000) | −0.000 (0.000) | ||
N | 70,160 | 38,059 | 70,160 | 38,059 |
0.288 | 0.191 | 0.206 | 0.249 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
2SLS, two–stage least squares; SC/ST, scheduled caste/scheduled tribe.
2SLS: effect of living in a household that owns a refrigerator on married women's work
Owns Refrigerator (single women) | 0.376*** (0.0132) | 0.352*** (0.0204) | 0.376*** (0.0132) | 0.352*** (0.0204) |
Owns air conditioner | 0.158*** (0.0190) | 0.223*** (0.0313) | 0.158*** (0.0190) | 0.223*** (0.0313) |
Owns color television | 0.121*** (0.00848) | 0.0999*** (0.0112) | 0.121*** (0.00848) | 0.0999*** (0.0112) |
495.87 | 193.8 | 495.87 | 193.8 | |
Owns refrigerator | −0.145*** (0.030) | −0.106* (0.058) | 0.082*** (0.026) | 0.010 (0.045) |
Age | 0.037*** (0.002) | 0.050*** (0.002) | 0.034*** (0.001) | 0.038*** (0.002) |
Age–squared | −0.000*** (0.000) | −0.001*** (0.000) | −0.000*** (0.000) | −0.001*** (0.000) |
Years of schooling | −0.006*** (0.001) | −0.009*** (0.001) | −0.002** (0.001) | −0.007*** (0.001) |
Rural | 0.285*** (0.014) | 0.072*** (0.026) | 0.010 (0.011) | −0.016 (0.020) |
Hindu | 0.053*** (0.009) | 0.089*** (0.016) | 0.026*** (0.006) | 0.012 (0.009) |
SC or ST | −0.065*** (0.008) | −0.079*** (0.010) | 0.113*** (0.008) | 0.136*** (0.009) |
Wealth index | −0.025** (0.010) | −0.012 (0.016) | −0.150*** (0.010) | −0.136*** (0.014) |
No. of children (household) | 0.004*** (0.002) | −0.000 (0.002) | −0.005*** (0.001) | −0.005*** (0.002) |
Distance to nearest town (km) | −0.001 (0.000) | 0.000 (0.000) | ||
Distance to road access (km) | 0.003* (0.002) | 0.002 (0.002) | ||
Phone access | −0.046*** (0.013) | 0.016 (0.010) | ||
Distance to railway station (km) | 0.000 (0.000) | 0.000 (0.000) | ||
Distance to market (km) | 0.003*** (0.001) | −0.002*** (0.001) | ||
Distance to secondary school (km) | 0.003*** (0.001) | 0.000 (0.001) | ||
Distance to higher secondary school (km) | 0.000 (0.000) | −0.000 (0.000) | ||
N | 70,160 | 38,059 | 70,160 | 38,059 |
0.288 | 0.191 | 0.206 | 0.249 |
Denotes significance at 10%.
Denotes significance at 5%.
Denotes significance at 1%. Standard errors are clustered at the district level.
2SLS, two–stage least squares; SC/ST, scheduled caste/scheduled tribe.
Table 9 presents the effect of washing machine ownership on married women's employment segmented by work inside and outside the household (Columns 1–2 and 3–4, respectively). We find that married women living in households that own a washing machine experience a decrease in the probability of being employed in work inside the household by 6.8 percentage points in the full sample (Column 1). Employment in work inside the household is reduced by 14.7 percentage points in the village sample (column 2). We find that married women's probability of employment in work outside the household increases by 15.4 and 8 percentage points in the full and village samples, respectively (columns 3 and 4).
Table 10 presents similar results of owning a refrigerator on married women's employment by different work types. The results suggest that ownership of a refrigerator decreases the probability of married women's participation in work inside the household by 14.5 percentage points in the full sample (Column 1). These results hold when we extend our analysis to the village sample, with married women's probability of being employed in work inside the household being reduced by 10.6 percentage points in the village sample (Column 2). Similar to the washing machine-related findings, refrigerator ownership is found to increase the probability of married women's employment in work outside the household by 8.2 percentage points in the full sample.
The results presented in Tables 9 and 10 are suggestive of appliances being used as a tool to assist women in altering their time commitments between employment in different work types. Appliances allow married women to reallocate their time to jobs that are of more value to them as both appliances are found to significantly alter work force decisions for married women. The results suggest that time-saving appliances allow women more freedom of choice, allowing them to select employment that is less home-based and potentially more lucrative and liberating.
Previous literature has stressed the importance of household capital in the form of time-saving technologies in altering women's and children's labor force and education decisions in other countries (Coen-Pirani et al., 2010; Tewari and Wang, 2016; Kerr, 2019). Our paper is the first to discuss the effects of time-saving appliance ownership on adult female and child outcomes in the Indian context.
Using detailed data from the IHDS and observing two time-saving appliances, we study the effect of appliance ownership on married women's employment, older children's employment, and school enrollment. We find that time-saving appliance ownership reduces the probability of married women's overall employment. Ownership of time-saving appliances may increase the relative productivity of married women in household production, and if these appliances are complements to household work, we would expect to observe an increase in household production and decrease in overall employment for married women. Given the strong patriarchal nature of the Indian society and the presence of traditional gender roles, it is expected that Indian married women perform most of the household chores in an Indian household. This is supported by the data provided by the OECD at:
However, when categorizing employment type by economic activity inside and outside the household separately, we find that married women decrease the probability of being employed inside the household and increase the probability of being employed outside the household when living in a household that owns a time-saving appliance. These results suggest that, even with India's Hindu culture, caste system, and reduction in overall female LFP, appliance ownership improves married women's outcomes.
Ownership of a refrigerator increases the probability of older children being enrolled in school and reduces the probability of them being employed. Living in a household that owns time-saving appliances increases a child's school enrollment by providing the child the opportunity to reallocate their time from household labor to schooling. The presence of a time-saving appliance lowers the opportunity cost of staying in school by directly reducing children's time allocated to household work.
Given the current and projected growth in the ownership of such time-saving durable household appliances in the Indian economy, our paper throws light on an important channel that affects both women and children outcomes. This is particularly relevant for researchers and policymakers who are interested in understanding the drivers of Indian married women's declining LFP, a phenomenon that is counterintuitive and the opposite of what is being observed in other developing and developed countries.