17Similarly, because households were so poor, almost all food (98 percent) was at-home consumption, so respondents were aware of the exact ingredients and quantities used. 7Ours is not the first study to suggest rice as a likely candidate for Giffen behavior. Dwyer and Lindsay (1984) propose (but do not test) this possibility for Singapore, and John-Ren Chen (1994) finds suggestive evidence of positively sloped demand for rice in Taiwan.
Private goods are goods with private ownership which means such goods are exclusive (Excludable) to its owner. Such goods require permission from the owner for consumption by other individuals. Goods that are considered inferior in developed countries may not be considered inferior in developing countries. For
example, the United States might consider simple fast foods as inferior goods, but developing countries, like India or Bangladesh or Bhutan, might not.
Inferior goods and consumer behavior
These examples are usually related to staple goods, which are items that make up a significant portion of a consumer’s budget. Inferior goods are those goods whose demand falls with a rise in the income of consumers. 40Most shopkeepers sold only grain, so most households could not have exchanged the vouchers for other foods.
- However, we cannot rule out that some noodles were made at home from flour but recorded as noodles, or that some consumers mistakenly reported purchased bread as wheat flour.
- A number of economists have suggested that shopping at large discount chains such as Walmart and rent-to-own establishments vastly represent a large percentage of goods referred to as “inferior”.
- Thus, splitting the data by the pre-intervention or initial share of caloric intake from consumption of the staple (Initial Staple Calorie Share, or ISCS) provides a more direct measure of whether a consumer or household is well off enough that they could, potentially, exhibit Giffen behavior.
- Giffen goods are special types of goods whose demand increases with the rise of price and decreases with the fall of the price.
- Second, the vouchers were distributed only incrementally (one-fifth each month) over the course of the intervention (though all vouchers remained valid until the end of the intervention).
These regional differences in preferences are primarily determined by geography, climate, and history, with wheat the dominant crop grown in Gansu and rice dominant in Hunan. Accordingly, we subsidized rice (only) in Hunan and wheat flour (only) in Gansu. In his textbook Principles of Economics, economist Alfred Marshall described Robert Giffen’s giffen goods example in india work in the context of bread rising in price because people lacked the income to buy meat. Randomly selected households in both provinces were given vouchers that subsidized the purchase of their respective staple foods. In the case of Giffen goods, the income effect can be substantial while the substitution effect is also impactful.
Further, there is no way to cut back consumption of these foods while maintaining protein intake; with meat, households can reduce consumption but switch to pulses as a less expensive source of protein. The merchants in these shops agreed to honor the vouchers in exchange for reimbursement and a payment for their participation. Households and merchants were told they were not permitted to exchange the vouchers for anything but the staple good, that there would be periodic auditing and accounting to make sure they were in compliance with the rules, and that any violations would result in their being removed from the study without any additional compensation. Households and merchants were explicitly told that selling the vouchers for cash or reselling rice or wheat bought with the vouchers would result in dismissal from the program.
Overview of Inferior Good
Thus, as theorized by Marshall and others, when faced with an increase in the price of the staple good, these households do, indeed, “consume more, and not less, of it” (Marshall 1895). While the theory suggests we should also exclude the wealthier households in the standard zone of consumption, unlike the threshold for segregating households that are too poor, it is unfortunately not possible to estimate the threshold for this region. Further, because our sample is drawn from the poorest households, there is no guarantee we even have any households in this zone. Therefore, we begin by taking the conservative approach of using only the threshold excluding the poorest. If our theory is correct, if anything, keeping the lower tail of the staple calorie share distribution will make it less likely we find Giffen behavior, since we are potentially including households with downward sloping demand among our potential Giffen consumers (we explore this possibility in Section IIIC). The consumer’s response to an increase in the price of the staple good will differ across the three regions of his indifference map.
Further, these goods, and the populations that exhibit Giffen behavior, meet some basic but common conditions that suggest this behavior may be widespread in the developing world. Thus, the absence of previously documented cases most likely results from inadequate data or empirical strategies rather than from their nonexistence. Unfortunately, classifying households or individuals directly by consumption zone is not possible. Not only is there no consensus on what constitutes a subsistence level of calories, but any such threshold would certainly vary widely by age, sex, height, weight, body fat, muscle composition, level of physical activity, health status, and a range of other factors. As a result, although we can compute caloric intake for each individual, identifying whether specific individuals are below, near, or above their subsistence level of caloric requirements is not possible. For the same reason, it is not possible to define these regions based on income or expenditure; individuals with different characteristics will require different amounts of expenditures or income to achieve nutritional sufficiency.
Giffen Behavior and Subsistence Consumption
Typically, when the price of a good increases, the demand for that good decreases. The good must either have a lack of close substitutes or the substitute good must have a higher cost than the good. Even if there is an increase in the price of the good, the current good should still be an attractive option for the consumer.
Selecting just the very poorest, or even aggregating over a broader set of households that includes both those in the calorie-deprived and subsistence zones, may not be sufficient. Consumers in the intermediate, subsistence range must be isolated in order to find such behavior. A Giffen good is a low income, non-luxury product that defies standard economic and consumer demand theory. Demand for Giffen goods rises when the price rises and falls when the price falls. In econometrics, this results in an upward-sloping demand curve, contrary to the fundamental laws of demand which create a downward sloping demand curve. When we restrict our sample to take these factors into consideration, we do find evidence of Giffen behavior, though the samples are smaller, precisely because most households do not conform to the conditions in Gansu.
Any such cutoffs would be imperfect, including some people who, because of high weight or activity levels, are unable to achieve maintenance nutrition with the specified income, and excluding others who have lower than expected nutritional (and thus income) needs because of small stature or low activity levels. Giffen goods are special types of goods whose demand increases with the rise of price and decreases with the fall of the price. A giffen good shows an upward-sloping demand curve and it generally violates the fundamental law of demand. 36Though in selecting sample sites, the authors personally visited only two of the counties in Gansu (Anding and Yuzhong). These counties, both with significant Muslim populations who traditionally consume primarily the home made bread mo, fit the pattern better, with 88 percent of all wheat consumption coming from flour, compared to 74 percent in the other three counties.
When the consumer is relatively wealthy, he will demand a bundle of goods in the standard zone. In this case, as illustrated in panel B of Figure 1, we expect the consumer to respond to an increase in the price of the staple good by consuming less of it. As wealth decreases, the consumer’s demand moves into the subsistence zone, and the consumer focuses more on maintaining caloric intake as his primary goal.
The safeguards discussed above were accompanied by monitoring and auditing to check for compliance. These audits did not discover any such cheating, and our survey personnel, who visited the households, did not discover evidence of cashing out. Finally, comparing voucher redemptions relative to estimated consumption and storage suggests that to the extent it occurred at all, cashing out could not have been significant or widespread (details in the online Appendix).
Consumers thus choose to decrease their consumption of inferior goods when real income increases, showing the inverse relation between inferior goods and the income effect. 26Ideally, we would use the data from each particular round to assess living standards rather than using only the preintervention data, since Giffen behavior depends on a consumer’s budget at the time he makes his decisions. However, expenditure in the round with the subsidy is obviously endogenous with respect to the subsidy; income would encounter enodgeneity as well (the increased consumption afforded by the subsidy might affect earnings). 22The broad conclusions of our analysis hold if we instead use, say, staple budget shares to parse the data (see the online Appendix). First, expenditure data are notoriously noisy, especially due to large but infrequent purchases such as durable goods that can skew budget shares. By contrast, the diets of households in our sample rarely contain more than three or four items and typically do not vary from day to day.
While this is unlikely to happen often in reality because the price of all the forms of the staple will be linked to the price of the raw ingredient (here, wheat), the unique structure of our subsidy did just that, subsidizing only the form of the staple prepared at home, and not the close substitutes purchased in stores. Among this group there is again statistically significant evidence of Giffen behavior, with a very large elasticity. A second prediction of the subsistence model we can explore is that, in response to an increase in the price of the staple good, consumers facing a subsistence constraint will consume not only more of that good, but also less of the fancy good, which we identified here as meat.
Income and substitution effects
Thus, the consumption of all other foods combined on average contributes only 10 percent of calories in Gansu, and 15 percent in Hunan. In Hunan, the greatest remaining share comes from meat, with 42 grams of consumption per person per day on average, comprising 7 percent of average caloric intake. By contrast, in Gansu meat consumption is much lower, averaging only 13 grams per person per day and contributing less than 1 percent of total caloric intake. This is likely due to the lower income levels in this province; pulses are often referred to as “poor man’s meat” because they are a cheaper source of protein (when combined with other foods typically eaten as staples).
For example, an employee, despite receiving a pay raise, might still prefer to buy her coffee from McDonald’s and not from Starbucks. This would be an example of personal preference, which does not conform to regular consumer behavior. 39Cheating, where shopkeepers do not provide the full subsidy to consumers (for example, those with poor math skills), effectively lowers the value of the subsidy, so the Giffen behavior we find would likely have been even stronger had such cheating not occurred. 34This result was unanticipated, since the northern provinces in our original paper (Jensen and Miller 2002), and our field test of the survey for the current study, revealed considerably more meat consumption in Gansu. Dwyer Jr. and Cotton M. Lindsay (1984) present a summary of the basic case against the potato version of the Giffen paradox.