THE CHOICE OF PAYMENT VEHICLE IN DISCRETE CHOICE EXPERIMENTS: LABOUR VERSUS MONEY

Abstract

There is a growing literature advocating for the use of non-monetary payment vehicles (PVs) in stated preference (SP) studies. It is typically argued that the use of a monetary PV underestimates willingness to contribute in subsistence-oriented communities where households often lack access to waged labour opportunities. Many studies use labour as a non-monetary PV which leads to estimates of willingness to work (WTW) in place of, or alongside, estimates of willingness to pay (WTP). This thesis comprises a series of chapters investigating issues around the use of labour as the PV. The empirical analysis is based on two discrete choice experiments - one concerns fortified flour in Kenya and one concerns water scarcity in India. Respondents in both studies were randomly assigned to treatments where the PV was either money or labour. In the Kenyan study, respondents were further randomly assigned to a treatment where choices were either hypothetical or consequential. Chapter 1 uses data from the Kenyan field study to investigate issues regarding monetisation of WTW. To obtain a monetary measure of welfare, which is a requirement in cost benefit analysis, estimates of WTW are often monetised ex post. A fundamental challenge, however, is the need to apply an opportunity cost of time (i.e. a rate for converting labour to money). Most studies use some proportion of the wage rate as a proxy for the opportunity cost of time but there is no consensus about the most appropriate conversion rate to use. Estimates of WTP (based on a monetary PV) are often used as the benchmark against which the performance of one or more conversion rates is evaluated. Previous (comparative) studies of PV effects are based on hypothetical SP studies. As a result, the benchmark against which monetised WTW is assessed is hypothetical WTP. The Kenyan study design allows a comparison of the performance of six commonly employed conversion rates using hypothetical WTP (which is standard in the literature) and consequential WTP (which is unique to this study) as the benchmark. The results in Chapter 1 indicate that the best performing rate when hypothetical WTP is used as the benchmark performs poorly when consequential WTP is used as the benchmark thus demonstrating the importance of the choice of benchmark. In Chapter 2, data from the Kenyan field study is used to test for differences in hypothetical bias (HB) between a monetary and a labour PV. A recurrent finding in the WTW literature is that monetised WTW exceeds WTP. One possible explanation is higher levels of HB for labour PVs. Chapter 2 investigates this claim and finds that hypothetical bias is 26 to 31 percentage points higher when a labour PV is used instead of money. Chapter 3 uses data from the Indian study to test for gender-based differences in the opportunity cost of time in the context of a patriarchal family structure. The results show that women, ceteris paribus, are WTW more than men while men are WTP more than women. The estimated shadow value of time is thus lower for women than for men which suggests that the choice of PV impacts not only welfare values in general but also the social importance attached to men and women in welfare valuation.

Date of Award	1 Aug 2021
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Daniel Rigby (Supervisor) & Prasenjit Banerjee (Supervisor)