Little is currently known about the effects of shocks to parental health on the allocation of children's time between alternative activities. Using longitudinal data from the Ethiopian Young Lives surveys of 2006 and 2009, we analyse the effect of health shocks on the amount of children's time spent in work, leisure and education. One key contribution of the paper is that we distinguish between child labour as defined by organisations such as the International Labour Organisation and other types of child work, such as light domestic chores. We find that paternal illness increases the time spent in income-generating work but maternal illness increases the time spent in domestic work. Moreover, maternal illness has a relatively large effect on daughters while paternal illness has a relatively large effect on sons. Overall, parental illness leads to large and significant increases in the amount of child labour.

Much of the existing literature on the economics of child labor assumes that child labor is synonymous with employ-

SUPPLEMENTARY MATERIALS S1. Models with Regional Ethnic Diversity as an InstrumentTable A1 includes estimates of the effect of neighbourhood ethnic diversity (nebdiv) on the child health outcomes when regional ethnic diversity (regdiv) is used as an instrument instead of zonal ethnic diversity (zondiv). The results in Table A1 are to be compared with those in Tables 2-3 of the main text. The estimated marginal effects in the model of immunisation uptake in Table A1 are very similar to those in Table 2. The estimated effect of nebdiv is slightly less precise than in Table 2, and is significant at only the five percent level, but our conclusions about the determinants of immunisation uptake are not affected by the change of instrument. The estimated effects on weight-for-height in Table A1 are very similar to those in Table 3, except that the effect of nebdiv is about twice as large when regdiv is used as an instrument. S2. The Validity of the Instruments for Neighbourhood DiversityThis section includes two sets of results pertaining to the validity of our instrumental variables (zondiv and regdiv). Firstly, Table A2 shows results from the first-stage regressions of nebdiv on the instrumental variables plus the other explanatory variables in Table 3 of the main text. This is to check that there is a significant association between nebdiv and the instruments, conditional on the other explanatory variables. Secondly, Table A3 shows results from models of immunisation uptake and weight-for-height that include an additional set of zone-level explanatory variables. This is to check that our exclusion restrictions are valid, i.e. that the use of a zone-level or region-level ethnic diversity measure as an instrument for nebdiv still produces significant associations between our health outcomes and nebdiv, even when we control for other zone-level characteristics. These characteristics are zonal wealth (a zone-level average of the household

Given the conventional wisdom that poverty and associated income shocks are the fundamental causes of child labour, from a policy perspective, there is a perception that social safety net programmes can play a vital role in reducing child labour. While there is extensive evidence that shows the prevalence of child labour is low among beneficiaries of conditional cash transfer programmes, the impact of workfare programs on child labour has been rarely investigated in the economics literature. This paper addresses the issue by evaluating the impact of the public works component of the Productive Safety Net Programme implemented in Ethiopia in 2005. The programme aims to help poor households to build assets and develop resilience to shocks through employment in public projects. Results from child fixed effects estimations show that children in programme beneficiary households are significantly less likely than their counterparts in non‐beneficiary households to be involved in child labour. The findings suggest that, if well targeted, even safety net programmes that do not primarily target child outcomes can be useful in addressing child labour problems.

The effect on economic outcomes of ethnic diversity remains debatable. Some view ethnic diversity as a deterrent of development, whereas others consider it a source of innovation and productivity, which can be translated into a higher level of development. This study aims to shed further light on the issue. Applying instrumental variables estimation to district‐level data, we find that ethnic diversity has a significant positive association with nighttime light intensity, which is used as a proxy for level of economic activity. The finding is robust to alternative specifications. Our finding suggests that centuries of interethnic contacts and coexistence may have helped ethnically diverse communities to gain experimental knowledge of the diverse beliefs and social practices and transition it into better economic outcomes.