Abstract
Spatial version of multivariate Fay–Herriot model is introduced and
small area predictor under this model is proposed. The mean squared
error (MSE) estimation of the proposed small area predictor is also
developed. The empirical performance of the proposed small area
predictor and the MSE estimator are evaluated through simulation
studies. The empirical results clearly show that the proposed small area
predictor outperforms the existing predictors. The proposed MSE
estimator tracks the actual value of MSE reasonably well with acceptable
coverage rate. An application to estimate the disparities in food and
nutrition intake from the 2011–12 Household Consumer Expenditure Survey
data collected by the national sample survey office of India is also
presented.