RESEARCH

Research Interests:  

Income distribution is how the national income is distributed amongst its population, either individuals or households. Important theoretical and policy concerns include the relationship between income distribution and income inequality and poverty. The gap between rich and poor and the number of people below the poverty line have both grown over the past two decades in the world. The increase is widespread, affecting three-quarters of OECD countries. The rise in inequality is generally due to the rich improving their incomes relative both to low- and middle-income people. Around one person in 10 in OECD countries had an income below half of the national median in 2005. The risk of poverty for older people has fallen, while poverty of young adults and families with children has risen. Work reduces poverty: child poverty is lower in countries where more mothers work.

Modeling income distribution as a mixture distribution provides a natural framework for the detection of homogeneous subgroups of population. In modeling cross-country distribution of per capita income, the mixture components can be interpreted as convergence clubs. One possible manifestation of the presence of convergence clubs is multiple modes in the cross-country distribution of per capita income with each mode corresponding to a convergence club. Kernel density estimator allows for a visual impression of the underlying distribution in order to detect multimodality.

A characteristic of most large surveys on income data is that some of the intended measurements are not available. This may occur through chance or because certain question are unanswered by particular groups of respondents (in income surveys for example rich people, are particularly unwilling to answer questions regarding income or earnings) or may not respond at all to a particular wave of a longitudinal survey. Understanding how income surveys work with missing data is an important issue and establishing mechanisms to track sources of missing data may help in understanding income data and the related empirical results on poverty, inequality and economic situation in general.

Multilevel models, or hierarchical models, are statistical models of parameters that vary at more than one level. The basic idea of hierarchical modeling is to think of the lowest-level units (smallest and most numerous) as organized into a hierarchy of successively higher-level units. For example, individuals reside in regions, regions are in nations. Hierarchical models are often applicable to modeling of data from complex surveys, because usually a clustered or multistage sample design is used when the population has a hierarchical structure. Different predictors may be relevant on different levels, like individual incomes and per capita GDP of a country and eventually their interactions.