Research Centre on Micro-Social Change (MISOC) through University of Essex

Principal Investigator: Man Yee Kan


The core activity of the Centre is to conduct an annual panel study (the British Household Panel Study) to look at social and economic behaviour at the household level. The study involves a nationally-representative sample of 5,000 households (some 10,000 individuals) in Britain.

Dr Kan is responsible for two projects under the MISOC grant:

  1. The distribution of financial savings, wealth and debts within married and cohabiting couples - in collaboration with Heather Laurie at University of Essex

The distribution of financial assets and wealth between married and cohabiting partners has rarely been explored. Past studies tend to adopt a household-based approach and assume that wealth is evenly distributed among members of the household. This project employs an individual-based approach and aims to investigate:

  • The likelihood of men and women having savings, investments and debts respectively, and in particular, the patterns of moving in and out of debts over time;
  • The types of savings, investments and debts held by each partner and whether there are any gender differences;
  • Whether the financial assets and debts are held jointly or solely, and the trends of joint/sole financial asset holdings over time;
  • The amount of wealth held by men and women at the individual level and factors influencing the accumulation of wealth over time.

The data used are from the British Household Panel Survey, in which additional information on financial assets and wealth was collected in 1995, 2000 and 2005. We will provide both cross-sectional information on trends across the ten year period and a longitudinal analysis of changes in wealth holdings over time in relation to other life course events, labour market participation patterns and changes in household composition.

II. Estimating measurement error in time use data – in collaboration with Stephen Pudney at University of Essex

This project aims to estimate measurement error in the two most common types of time use data: "stylised" questionnaire-based data and diary-based data, and develop methods to correct the biases induced by these measurement errors. The following key aspects of our project:

  • The reporting of time spent on household work is prone to be biased due to social desirability factors. We will analyse stylised estimates and diary-based estimates collected from the same respondents to find out the nature and size of measurement the error biases in these estimates. The data sources are the Home On-Line Survey (HoL), 1999-2001, which contains both types of estimates and the British Household Panel Survey (BHPS), 1991-2005, a major panel survey in the UK which collected stylised housework and paid work time estimates annually.
  • There have been debates among academics and policy makers about how to collect data concerning the length of workweek accurately and efficiently. We will examine and compare market work time estimates in HoL (which collected weekly diaries and stylised weekly work hours from the same respondents), BHPS (which collected stylised work hours annually), Labour Force Surveys (which collected stylised weekly work hours) and the ONS UK Time Use Survey (UKTUS), 2000 (which collected stylised work hours and "work grid" weekly work time estimates and one-day diaries from the same respondents). We will evaluate the accuracy of market work time estimates collected by the traditional diary approach, the survey interview method and the work-grid diary method.
  • Although diary time use data are generally preferred by researchers, many people may find keeping a diary burdensome, and hence diary data are likely to contain certain selection biases. We will investigate whether selection biases exist in data collected by the weekly diary method, the one day diary method and the work-grid diary method respectively. The data sources are HoL and UKTUS.

Funding Provider

Visit the Economic and Social Research Council website

Economic and Social Research Council

Duration of Funding: 5 years