Final Report Summary - MICROCONLAB (Microeconometrics of Consumer and Labour Supply Behaviour)
There were two key objectives of the research programme for the ERC MicroConLab project. The first was to extend the analysis of consumer behaviour and nonparametric revealed preference to allow for preference heterogeneity, multiple goods and taste change. The second objective was to examine consumption and labour supply behaviour in a dynamic life-cycle environment.
In terms of the first objective we have developed methods that allow the imposition of shape restrictions that derive from revealed preference integrability restrictions and that deliver counterfactual policy scenarios in the nonparametric analysis of consumer behaviour. There have been two broad major advances as a result of this research: First, to utilise inequality restrictions deriving from revealed preference theory to improve demand estimation and prediction. Second, to relax restrictions on unobserved heterogeneity in consumer demand functions.
In terms of the second objective the technical achievements have been to develop a robust model of non-separabilities between consumption and family labour supply, to develop methods for nonlinear dynamics, and to allow for heterogeneity and constraints on the behavioural responses of consumption and family labour supply behaviour to policy reforms and other changes in the economic environment. There have been two broad major advances in relating to new models and new estimates that explore the importance of non-separabilities, dynamics and heterogeneity.
The development and application of the new quantile-based panel data framework to study the nature of income persistence and the transmission of income shocks to consumption has shown that the approach is not only feasible but has changed the way we understand income and consumer dynamics. The results show nonlinear persistence and conditional skewness to be key features of the earnings process. This research also showed that the impact of earnings shocks varies substantially across earnings histories, and that this nonlinearity drives heterogeneous consumption responses. The transmission of shocks has been found to vary systematically with assets. The estimated quantile Markovian permanent-transitory model of income reveals significant asymmetric persistence of earnings and income shocks.
The Norwegian Register data has enabled a fine granular examination of household behaviour and we have shown the importance of different models of wage dynamics in determining consumption over the life-cycle using this new registrar data on family earnings and savings. This has established a large degree of transmission of both income and house price shocks to household consumption. Our results point to the importance to allow for correlated changes in local labor market conditions and house prices to understand consumption behaviour. Accounting for generational links is found to be key to understand how households respond to house price changes.
Work on family labour supply and consumption intertemporal interactions has shown that with a flexible life cycle model with self-insurance through savings, endogenous family labor supply, non-separable preferences, government transfers and progressive taxation, there is little evidence of "missing" insurance explaining consumption movements in response to wage shocks. We find a sizable Frisch labor supply elasticities for both husband and wife, but show that the implied Marshallian elasticities are much smaller due to strong wealth effects. Adding in time use data our results highlight that non-work times in couples are complementary goods when used for companionship whereas for time use with children, at least young children, they are substitutes. Work on dynamics of labour supply and earnings, have shown that work experience is strongly complementary with education. We estimate lower returns for the lower educated and for those in part-time work. This work has established that human capital investments in work explain around 60% of the gender wage gap.
Research on nonparametric demand showed reveals systematic variation in price responsiveness across the income distribution. It showed the key importance of allowing for Berkson measurement error in demand models. In these nonlinear models, Berkson errors are not innocuous and require careful treatment. Importantly, these kinds of models are of growing importance for applied researchers, given (i) the recognition of the importance of heterogeneity within populations, and of uncovering and understanding these patterns of heterogeneity, and (ii) the availability of large micro datasets and computational power that allow to estimate more flexible specifications.
In terms of the first objective we have developed methods that allow the imposition of shape restrictions that derive from revealed preference integrability restrictions and that deliver counterfactual policy scenarios in the nonparametric analysis of consumer behaviour. There have been two broad major advances as a result of this research: First, to utilise inequality restrictions deriving from revealed preference theory to improve demand estimation and prediction. Second, to relax restrictions on unobserved heterogeneity in consumer demand functions.
In terms of the second objective the technical achievements have been to develop a robust model of non-separabilities between consumption and family labour supply, to develop methods for nonlinear dynamics, and to allow for heterogeneity and constraints on the behavioural responses of consumption and family labour supply behaviour to policy reforms and other changes in the economic environment. There have been two broad major advances in relating to new models and new estimates that explore the importance of non-separabilities, dynamics and heterogeneity.
The development and application of the new quantile-based panel data framework to study the nature of income persistence and the transmission of income shocks to consumption has shown that the approach is not only feasible but has changed the way we understand income and consumer dynamics. The results show nonlinear persistence and conditional skewness to be key features of the earnings process. This research also showed that the impact of earnings shocks varies substantially across earnings histories, and that this nonlinearity drives heterogeneous consumption responses. The transmission of shocks has been found to vary systematically with assets. The estimated quantile Markovian permanent-transitory model of income reveals significant asymmetric persistence of earnings and income shocks.
The Norwegian Register data has enabled a fine granular examination of household behaviour and we have shown the importance of different models of wage dynamics in determining consumption over the life-cycle using this new registrar data on family earnings and savings. This has established a large degree of transmission of both income and house price shocks to household consumption. Our results point to the importance to allow for correlated changes in local labor market conditions and house prices to understand consumption behaviour. Accounting for generational links is found to be key to understand how households respond to house price changes.
Work on family labour supply and consumption intertemporal interactions has shown that with a flexible life cycle model with self-insurance through savings, endogenous family labor supply, non-separable preferences, government transfers and progressive taxation, there is little evidence of "missing" insurance explaining consumption movements in response to wage shocks. We find a sizable Frisch labor supply elasticities for both husband and wife, but show that the implied Marshallian elasticities are much smaller due to strong wealth effects. Adding in time use data our results highlight that non-work times in couples are complementary goods when used for companionship whereas for time use with children, at least young children, they are substitutes. Work on dynamics of labour supply and earnings, have shown that work experience is strongly complementary with education. We estimate lower returns for the lower educated and for those in part-time work. This work has established that human capital investments in work explain around 60% of the gender wage gap.
Research on nonparametric demand showed reveals systematic variation in price responsiveness across the income distribution. It showed the key importance of allowing for Berkson measurement error in demand models. In these nonlinear models, Berkson errors are not innocuous and require careful treatment. Importantly, these kinds of models are of growing importance for applied researchers, given (i) the recognition of the importance of heterogeneity within populations, and of uncovering and understanding these patterns of heterogeneity, and (ii) the availability of large micro datasets and computational power that allow to estimate more flexible specifications.