Economic preferences across states in India4 min read . Updated: 10 Jul 2018, 10:05 PM IST
Understanding why people in some states are more willing to take risks or trust others is important for analysing economic outcomes
Recently, behavioural economics researchers at the University of Bonn, University of Munich and University of Pittsburgh made publicly available a global dataset (goo.gl/1xqWT6) on economic preferences that surveyed nearly 80,000 individuals in 76 countries, including India. The Global Preferences Survey (GPS) (goo.gl/rE5Xdn) is the first-of-its-kind survey dataset that made use of economic experiments to measure preferences related to trust, altruism, reciprocity, risk-taking, and time preference. Armin Falk and his colleagues have, through this dataset, attempted to measure behaviour and preferences that economists were previously unable to.
Finding a precise measure of such preferences can be difficult and plagued with challenges. Survey measures may not capture them consistently, and experiments that do can often be difficult to replicate on larger scales. However, recent work in experimental economics has pointed toward the significance of such preferences in driving how individuals make decisions. There have also been studies that suggest there are significant cross-cultural variations (goo.gl/frHP47) in these preferences, particularly related to the intersection of social norms and cultural beliefs. To the extent that preferences related to trust, risk, and time preference exist and are reasonably stable, it is important to understand how these measures are related to economic outcomes, such as incomes.
Cross-country analysis recently published (goo.gl/vGmsot) by researchers in the Quarterly Journal Of Economics shows patience (measured using the preference for early vs. delayed payments) is significantly associated with higher per capita incomes globally. Other measures such as risk taking (measured using the preference for “lottery" type payments vs sure payments) were negatively associated with economic output. Other preferences were defined on trust (measured as the amount returned in an investment game), negative reciprocity (the preference for punishing deviations from a norm of giving), and altruism (the amount donated to a charitable organization). The authors suggest that such economic preferences could even explain labour market choices, savings decisions, and pro-social (or giving) behaviour. What we know little about, however, is how such preferences vary within a country such as India, where cultural diversity is much higher.
The GPS is nationally representative, in that it covers preferences for samples in 24 states/Union territories, sampling a total of 2,539 individuals across India. To keep the analysis tractable, simple statistical correlations are presented here to explore associations and contrast them with findings of the global study. Patience, as with the global study, is weakly positively correlated with state-level per capita incomes, and negatively with poverty levels. Interestingly, the strongest correlations came about with measures of negative reciprocity, indicating that punishing errant behaviour could be a significant aspect of our economic preferences. Negative reciprocity was positively associated with per capita incomes (similar to global estimates) and negatively with poverty counts. The association of risk taking with economic outcomes was opposite for Indian states; a higher risk-taking preference was weakly and positively correlated with per capita incomes. Unlike the global estimates, trust (measured using an investment game) was not significantly associated with economic outcomes, but only weakly (negatively) associated with poverty count for India.
There is considerable variation within India as well. States that ranked at the bottom for the patience preference (most prefer immediate rewards) include Goa, Uttar Pradesh, Odisha, and Gujarat. Individuals in Tamil Nadu and Kerala had the highest patience preference. States with a strong preference for risk taking included Assam, Karnataka and Chandigarh, while the most risk averse were Bihar and Jharkhand. States classified as risk averse also had the highest amounts of positive reciprocity—Jharkhand and Bihar. Those with the least amount of positive reciprocity were Chhattisgarh, Madhya Pradesh and West Bengal. States that had the strongest preference for negative reciprocity were Karnataka, Assam and Gujarat, while those that had weak punishment preferences included Bihar, Madhya Pradesh and Jharkhand. This appears to be significantly associated with altruism, as individuals in Bihar, Jharkhand and Karnataka gave the most, while individuals in Madhya Pradesh, Odisha and Uttar Pradesh gave the least.
The most interesting regional disparity comes when we study preferences related to trust, where there appears to be a stark contrast between states in the south (that trust more) and those in the north (which trust much less). But wait, there’s more. Preferences between northern (Chandigarh, Delhi, Haryana, Himachal Pradesh, Jammu and Kashmir, Punjab, Uttar Pradesh and Uttarakhand) and southern (Tamil Nadu, Andhra Pradesh, Kerala, Karnataka and Puducherry) states and Union territories differ significantly. On average, patience preferences among individuals in the south were nearly four times as much as in the north, while risk taking was also much higher in the south. In contrast, trust levels were much lower (and negative) in north India compared to close to zero for south India. This suggests that cultural differences between south and north India could also be strongly associated with such preferences, and therefore contribute to economic outcomes.
Datasets such as the GPS help in shaping our understanding of factors beyond what is observed (such as gender, years of education, or caste) and their influence on economic outcomes. The association of such preferences with other culture-specific characteristics (such as volunteering or donating to social causes) is important for validation. At the same time, measures of such preferences may need to be modified or contextualized to better reflect local cultural beliefs. For example, risk taking as a measure was not (positively) associated with economic outcomes in India as it was globally. Although the links between entrepreneurship and risk taking are mixed, one would expect, for example, a sample of Gujarati entrepreneurs to be more risk taking than others. Such precision in measuring preferences can only be accomplished through robust extension and further modification of these experiments.
Anirudh Tagat is at the department of economics at Monk Prayogshala, Mumbai.
Comments are welcome at firstname.lastname@example.org