undistributed income multiplies particularly when rapid economic growth offers enormous opportunities for capital accumulation through speculation in stocks, land and houses.
Furthermore, using household (which can make up of related or unrelated members) as the statistical unit in the HIS also raise some problems. First, while generally intra-household transfers are not considered as income, the salaries received by live-in servants and payment made by boarders to the household are taken into account as income. The HIS acknowledges that such cases do artifically inflated household incomes and thus affect inequality measures. Second, the size of households can also have an important bearing on income and measures of income inequality.
However, the biggest problem to my mind regarding income data in Malaysia arises from the fact that detailed statistics on income have never been published by the government. Up to today, detailed data on income, especially distribution of income, have largely remained sulit (government secrets). The only income statistics available are those which have been published in the five -year plan documents and their mid-term review plans. But these have appeared only in a summary form and the information available in recent documents have been very limited. Hence, researchers on income distribution see little of what is available. This is indeed a very sad situation. While the government acknowledges the importance of narrowing income inequalities in its development efforts, there is virtually no available data on this most significant policy issue to enable independent researchers to provide alternative views on the issue and suggest solutions to the existing problems. The cost related to this unfortunate situation was well summarized by Charles Hirschman (n.d) as follows:
