1. Shamim Mondal: Does this account for perhaps lack of formal credit markets, which might be driving excess savings, and not consumption growth?
Areendam Chanda: We will be spending time during the presentation discussing these issues.
2. Anonymous Attendee: A general question. Can you please share resources or explain on how to use sample weights when only subset of the sample is used? for instance in this study sample subset is HOH age 24 to 69.
Areendam Chanda: We don’t do anything unusual. We re-weight the weights of the applicable households. If you have any specific suggestions, we will be glad to hear.
3. Shamim Mondal: What is on the y-axis? log consumption per capita?
Neha Bairoliya: We have normalized consumption expenditure relative to age 24. So we are looking at growth in expenditure in both countries with respect to age 24. So. for example, In the US at age 50 there is a 50% growth in total consumption expenditure.
4. Gunjan Kumari: Why do you think that non-durable consumption which means food consumption will increase over the life cycle? or Non-durable include something else? Low elasticity of food consumption is well recognized in literature. Your results are justifying that prevalent hypothesis.
Areendam Chanda: You are right that food occupies a large fraction of non-durable spending, but at the end of the day, it is also only half of non-durables.
5. Yehuda Klein: How do single adult families differ?
Areendam Chanda: We havent looked at that. I cant imagine there are many Indian households that fit that description.
6. Nikhil: What if the number of children, and family growth is endogenous to explaining consumption? would scaling create measurement error in consumption if that is the case?
Areendam Chanda: This is more of an ’accounting exercise’ and not a causal exercise. It is natural to expect family to size to change over the life cycle.
7. Gunjan Kumari: I would like to know if you have done this analysis for different level of consumption quartile. because with increasing Income, we expect consumption of poor to increase but not for the rich.
Areendam Chanda: No, we did not do consumption quartiles since the same household can be in different quartiles based on their life-cycle. It can be done though. We do look at occupations though.
8. Anonymous Attendee: What is the distribution of family size in the two countries?
Areendam Chanda: That was presented in the second figure. Its there in the paper.
9. Abhishek Tripathy: Do you also explore heterogeneity based on the household’s susceptibility to idiosyncratic shocks?
Areendam Chanda: Good question. No, not specifically. But we look at occupations. Hopefully, that partly addresses your concern.
10. Gautam Mazumdar: As the consumption pattern across India in both cases rural and urban classified in their different age groups. Do we have results and pattern on Gender specific consumption pattern also?
Areendam Chanda: The expenditures are recorded at the household level. So it would be difficult to do gender specific consumption. That would have more to do with intra-household allocations. Its an important issue but one that we dont do here (and do not have the right data).
11. Abhinav Narayanan: What is the motivation behind comparing India and US? Is it only data availalbilty and the wide usage of PSID to understand these questions? 2. The statistics are mostly cross-sectional that can be estimated using NSS data as well. So whats the value addition of using the CPHS database which is unique because of its panel structure?
Areendam Chanda: The macro consumption literature is very well acquainted with PSID, and has developed around it . So this is a useful point of reference for those working in that area. NSS data is not available recently. Income is poorly measured, which is important for our derivations of savings and also Neha is now presenting some results from the panel aspects.
12. Shamim Mondal: Intentions are binary? So it is a linear probability model, rather than say a logit?
Areendam Chanda: Yes these are OLS estimates.
13. Anonymous Attendee: Hi, if I understand correctly, it’s not the same household head being tracked from age 24 onwards, but a collection of different households. So HH age 70 would exhibit a different behaviour (since they have been around for longer, older generation is more careful etc). The younger household could be endogenously more risky, so do you control for each household?
Areendam Chanda: You are right. In these kind of profiles in the research , one cannot track the same household. Riskiness is definitely a possibility but that may not necessarily describe US households vs Indian households. US households can afford more risk due to easier access to credit markets, but we havent specifically analysed that.
14. Anonymous Attendee: is there any variable that captures savings amount in CPHS
Areendam Chanda: No. We construct our own from their data.
15. Shamim Mondal: One possible thing to explore (probably already done, in that case please ignore): lack of health infrastructure and out of pocket expenses hurt. And perhaps the households are preparing for that. The expenditure really goes up for Indians at later ages compared to US.
Areendam Chanda: That is certainly a possibility. It does not go up nearly as much as that of US households.
16. Jose Antony: If it is projected forward, should not purchase of producing capital like tractors or capital result in an increase of income and savings. The findings - are they counter inuitive in that regard?
Areendam Chanda: It is the intention to purchase.
17. Subrata Mukherjee: Does your period of analysis of the longitudinal CPHS data include pre and during pandemic period? If Yes, then what changes you observed with regard to purchase of major and minor durables and their impacts on household savings? I am asking this question because India experienced a surge in domestic household savings during the pandemic period, especially last year.
Areendam Chanda: Everything is pre-pandemic. We stop in September 2019.
18. Shamim Mondal: In absolute terms, perhaps, but in relative terms is the out-of-pocket health expenses higher for US households compared to Indians?
Areendam Chanda: Much more - If you see the summary statistics.
19. Nikhil: Few questions (a) shocks in employment during this period 2015-19 could affect the consumption profile. Can these results be specific to the period? (b) How do consumer durables in US compare with that of India? Is India special in its treatement of durables? Also, should I expect a life cycle pattern even in these lumpy investments?
Areendam Chanda: There is a lot of literature on US durables- they are able to smooth much better.
20. Satadru Das: Can the identification of savings for lumpy investment be endogenous? In the sense that its not intention to buy lumping investment that increase savings rate in the period before but households are able to buy lumpy investments when they are able to increase their savings rate.
Areendam Chanda: The monotonic increase in savings rates over the entire life cycle suggests that this is not just opportunistic savings
21. Jose Antony: The intention to purchase is a strong proxy for actual purchase/capital investment hence the proposition holds. It seems counter intuitive in a some aspects.
Areendam Chanda: You make a good point about increases in income. We have not tracked that part. The savings could be more ambiguous. Higher income would allow both higher consumption and higher savings levels. Not so sure about savings rates?
22. Jose Antony: Would be interesting to look at different scenarios and project what could have happened for an article for a general audience. Assuming constraining / liberating factors are identified.
Areendam Chanda: Thank you for the suggestion.
23. Preeti Pushpalata Zanwar: I am not sure if this was talked about, what about financing of home, is it from savings in India or something else?
Areendam Chanda: Yes, that is our main idea- that homes are being financed by savings.