Methods
This study conducts a series of discrete-time event-history analyses of demographic responses to changes in local rice price and household landholding. Using a series of logistic regression models, the model relates an individual's probability of experiencing a given demographic event—death, marital birth, first marriage, or out-migration—in the interval of the next one year following the beginning of the interval.
Because the data for our analyses are constructed with a person-year as the unit of observation, individuals are likely to contribute more than one observation. This built-in interdependence among observations can affect the standard errors in multivariate analyses.’5 Thus, to take the effects of inter-correlation among observations obtained from the same individuals into account, we estimate the logistic regression with robust standard errors based on Huber's formula (Huber 1967).16
4.