There are two main arguments in this article. First, if wages and empl
oyment are to be used as indicators of changing levels of rural povert
y they need to be complemented by micro and meso level studies of how
increases or declines in wages and employment are distributed among in
dividuals and households. Secondly, if the nature of the relationship
between employer and labourer is to be understood, aggregates such as
'casual' labour need to be unravelled. Evidence from a study of two sm
all localities in rural West Bengal between 1991 and 1993 suggests tha
t the poorest workers receive the lowest remuneration across a range o
f informal contracts, including daily time-rate, piece rate, seasonal
beck-and-call and migrant labour arrangements. Levels of remuneration
are also determined by locally specific ideologies of gender and socia
l rank and by party politics.