完善的住宅計畫與政策，其關鍵乃在於政府能否精準的推估住宅需求量，並且依此需求數量提供正確的住宅資訊作為介入市場與否及程度的依據。而住宅交易是市場供需雙方均衡的結果，錯誤與不當的市場干預及預期，將會導致住宅市場的扭曲，造成高房價、高空屋率、低住宅品質等失衡現象，而且更將招致住宅計畫與政策的失敗。因此正確而且實際的預測市場的住宅需求，是本文最重要的目的。本文以「戶長率」為技術工具，並且配合方法回顧，比較不同的預測方法的優劣點，乃以住戶組成率方法預測住戶形成數量，建立「住戶形成」預估模式並且進而預測「住宅需求」量。台灣地區有關住宅需求的相關研究，多依個體經濟構建行為模式，討論個體住宅服務需求水準；本文乃依總體經濟建構「住戶形成」及「住戶消失」預測模式，並且進一步推估「住屋單元」需求量及市場「住宅需求量。本文以新竹市縣為例，修正戶長率觀念並且利用住戶組成率以1999年為基礎，預測2000年之住戶形成數，並且依空間別、住戶規模別詳細說明其預測結果。

The completion and performance of housing policy and planning is due to the precise forecast, the basis which determines the degree government-intervention, on the quantity of housing demand from the government importantly.The housing transaction is the equilibrium between demand and supply of housing market. The wromg and inadequate intervention and anticipation on housing will distort the market, such as high housing-value, high vacancy-rate, and low housing-quality, which will destroy the structure of housing policy and planning. Precise prediction on housing demand is the extra purpose of this paper. Household membership-rate method, after reviewing on the relative forecast methods on household formation, is the most adequate one to model housing demand in this paper. The most relative researches on housing demand in Taiwan area were to establish the behavior models, discussion on the level of housing service, on the basis of microeconomic theory. In this paper, it is attempt to predict the amount of household formation as the basis of housing demand by macroeconomics. The empirical study, Hsin-Chu County, shows the results of the prediction on the amount of household formation, based on the census data of 1999, by household membership rate method in 2000.