Figure 2: Interpret this figure.
Figure 2: Dupas (2010)
• Short-run subsidies and long-run adopBon of new health products: evidence from a field experiment • Field experiment for Olyset long-lasBng insecBcide-treated bed net (LLIN) • Two phases: • Phase 1: • Subsidy levels for Olyset LLINs randomly assigned across households. • Households had three months to acquire the LLIN at the subsidized price. • Prices varied from $0 to $3.80. • Phase 2: • A year later, all households in four villages were given a second opportunity to acquire an Olyset LLIN, but this Bme everyone faced the same price ($2.30). • Key quesBons asked: • How does the subsidy affect long-run investment in this technology? • In parBcular: are there learning effects and/or anchoring effects? Figure 2: InterpretaGon
• Figure 3 presents two averages for each Phase 1 price group: the average willingness to pay for a bednet declared at baseline, before households had received the first voucher; and the average willingness to pay declared at the follow-up. • Gaining access to a free or highly subsidized LLIN in the first year increases households reported willingness to pay for an LLIN two months later. • This suggests the presence of a learning effect which dominates any potenBal anchoring effect. Table 2: Interpret this table.
Table 2: Dupas (2010) – same paper!
• Examines learning in more detail: looks at social network effects. • The higher the proporBon of neighbors who received the high subsidy, the more likely the household is to have redeemed the voucher and purchased the LLIN. • When looking at the results using the within 500m radius definiBon of social networks: if all of a household’s received the maximum subsidy, the probability of redeeming the voucher increases by 21.5 percentage points. Table 3: Interpret this table.
Table 3: Atkin (2015)
• “The caloric cost of culture: evidence from Indian migrants.” • Based on noBon that there are substanBal and persistent differences in food preferences across social groups. • QuesBon: can such food cultures constrain caloric intake? • Key findings: • Migrants consume fewer calories than otherwise similar locals. • Migrants bring their own origin-state food preferences with them. • The gap in caloric intake is related to suitability of the migraBon desBnaBon to migrants’ origin-state preferences. Table 3: InterpretaGon
• Column 2: compared to other households in the village, migrant households have 6.2% higher per capita expenditure, 4.5% higher per-capita food expenditure, and consume 1.3% more calories per person. • Last three rows: do migrants face the same prices? • Migrants pay 0.3% more than non-migrants in the same village. • But controlling for food expenditure, there is no difference. • Column 3: migrant and non-migrant households seem more similar if the wife moved for marriage sample. While migrant households sBll have higher expenditures than non-migrants, the difference declines by a third. Online Response QuesGon #1
• “If all workers were paid piece rates then uncondiBonal transfers to poor households would likely not affect worker producBvity.” OR3, Q1: Background
• Foster and Rosenzweig, 1994. • Key quesBon: how do payment schemes (Bme-wage or piece rate) affect effort levels? • Measure effort: make use of the biological balance equaBon in which calorie intake and energy expenditure determine weight (body mass) change. • Key findings: • Time-wage payment schemes and share-tenancy contracts reduce effort compared to piece-rate payment schemes and on-farm employment. • The same workers consume more calories under a piece-rate payment scheme or in on-farm employment than when employed for Bme wages. OR3, Q1: Answer
FALSE. • Piece rate workers have a direct incenBve to increase their producBvity. So they may use the uncondiBonal cash transfer to increase their caloric intake, and be more producBve. • Time-wage earners, on the other hand, would not have an incenBve to invest more in caloric intake, since producBvity would not affect their wages. Online Response QuesGon #2
• “A recent arBcle suggested that the introducBon of cash transfers in remote villages in Mexico led to deforestaBon. This result is consistent with the findings of Cunha et al that cash transfers tend to affect the price of locally produced goods. “ OR 2, Q2
• Cunha, De Giorgi, Jayachandran, 2011 • “The Price Effects of Cash Versus In-Kind Transfers” • Compares how cash and in-kind transfers affect local prices. • Both can affect demand for normal goods, but only in-kind transfers also increase supply. • In-kind transfers should lead to lower prices than cash transfers, which helps consumers at the expense of local producers. OR2, Q2: Answer
True. • Cunha et al. have shown that cash transfers increase the price of local goods (due to a shio in demand). • In response to higher prices, producers will want to produce more. • If deforestaBon is a side-effect of increased producBon, then this will lead to more deforestaBon. Online Response #3
• “One disadvantage of in kind transfers targeted to poor households is that relaBve to cash transfers, a large fracBon of the benefit may accrue to beper off households.” OR2, Q3: Answer
• Cunha et al. (2011) again. • “In-kind transfers should lead to lower prices than cash transfers, which helps consumers at the expense of local producers.” • The answer depends on who is rich and who is poor. • If the consumers are the worse-off (which we assumed in class), then the statement is false. The benefit accrues to worse-off households. • If the producers are the worse-off (which seems to be suggested in the paper), then the statement is true. The benefit accrues to the beper off households. OR 1, Q1 • If Jensen had allowed people to buy or sell vouchers then he would have been less likely to find that households receiving a voucher that provides discounted rice end up consuming less rice.
Background • Sec/on: Price effects of consump/on and health. • Jensen and Miller (2008): Giffen Behavior and Subsistence Consump/on. – Subsidizing prices of dietary staples for extremely poor households in two provinces in China: find strong evidence for Giffen behavior for rice in one of the provinces. • Giffen good: demand decreases when price decreases. • Inferior good: demand decreases when wealth increases. Answer • FALSE. – If Jensen had allowed people to buy or sell vouchers, then if individuals sell the voucher the program would essen/ally have had just an income effect. – A decrease in quan/ty consumed as a result of the program would then just imply that the good is an inferior good. – Inferior goods are more common than Giffen goods. – So it would be easier to find what appears to be a Giffen good (but in fact is just an inferior good). OR 5, Q3 • If households are especially poor they may not migrate, even given proper incen/ves, because the risk of a failed migra/on is too high.
Background • Sec/on: Adop/on of Health Technologies I. • Bryan, Chowdhury, Mobarak (2014): Underinvestment in Profitable Technology: the case of seasonal investment – Randomly assign an $8.50 incen/ve to households in rural Bangladesh to temporarily out-migrate during the lean season.
Answer • FALSE. – Mobarak finds: "households that are close to subsistence - on whom experimen/ng with a new ac/vity imposes the biggest risk - start with lower migra/on rates, but are the most responsive to our interven/on" (pp. 1674).