As population in advanced countries ages, the welfare of elderly individuals and their ability to maintain consumption and a healthy life after retirement become important policy issues. Due to population aging, demand for long-term care is expected to rise. However, the costs of pensions, healthcare and other elderly support programs pose serious concerns for public budgets. Understanding the timing of retirement, consumption and saving decisions in later stages of life is of the utmost importance for policymakers. In this project, we aim to understand the unequal effects of pension and health policies on men and women building on a collective choice model in which household's decisions are the result of bargaining between partners.
The original contribution of this research will be a theoretical model and an empirical application to Italy.
We will develop a simple model of collective choice to describe households' consumption and saving decisions around retirement. The purpose is to understand how demographic and socio-economic characteristics affect these decisions. In the model, we will consider individuals who differ in earned-income, survival probabilities and bargaining power. Solving the model for singles and couples, will give us testable implications on consumption and saving behaviour of couples as compared to singles. Due to women having on average lower earned-income, higher life expectancy and higher preferences for the bequest motive (to be verified empirically), if the bargaining power within the couple weights men's preferences relative more than women's, we would expect (i) widows' consumption to be lower than widowers' and lower than single women and (ii) single women to save more than couples and couples to save more than single men.
The next step in the project will be an empirical investigation of Italian household consumption and saving decisions using the SHIW's Bank of Italy data set. The econometric analysis requires having a proxy for the partners' relative bargaining power. Alternatively, this can be estimated following the approach proposed by Browning et al. (2013) and simplified by Lewel and Pendakur (2008). This approach obtains an estimate of the relative bargaining power of household members from revealed preferences; the idea is to look for the weights in the household utility function that rationalize the observed saving behavior as efficient. Burkard (2017) and Butikoefer et al. (2011) have used this approach to study the allocation of consumption in elderly households. In this project, we will follow Lewel and Pendakur's (2008) methodology to estimate the relative bargaining power of two-member households in saving decisions. To this purpose, we need to modify their analysis to allow for intertemporal decisions. Once we will have adapted the methodology to our purposes, we will test the results of our model on Italian data. To identify each partner's utility function, we will use singles and widows/ers.
An important innovative part will be to consider the allocation of consumption expenditures to long-term care and how these expenditures affect savings.
We expect our analysis to be of some help in designing policy-initiatives to address inequality in leaving standard among the elderly.
The expected output of the project are at least two articles to be submitted for publication in internationally peer reviewed academic journal.
References
Browning, M., Chiappori, P.A., and Lewbel, A. (2013). Estimating consumption economies of scale, adult equivalence scales, and household bargaining power. The Review of Economic Studies 80: 1267-1303.
Burkard, D. (2017). Allocation of expenditures in elderly households and the cost of widowhood. Swiss Society of Economics and Statistics, 153 (4):371-401.
Butikoefer, A., A. Lewbel and S. Seitz (2011). Health and retirement effects in a collective consumption model of older households. Boston College Working Papers in Economics 767, http://fmwww.bc.edu/EC-P/wp767.pdf
Lewel, A. and k. Pendakur (2008). Estimation of collective household models with Engel curves. Journal of Econometrics, 147: 350-358.