Housing Starts - Empirical Study

Senior Seminar

Housing Starts: an empirical study

Introduction

The housing sector is of importance in virtually any economy. The housing market is not a market in a classical sense but a series of overlapping submarkets differentiated by location, age of dwelling, type of tenure, age and quality. It is a basic necessity all over the world. First, housing is an essential consumption element that takes up a large share of the typical household's budget. Second, housing is the dominant asset in many household's portfolio. Third, housing production and maintenance constitute an important segment of the economy's production sector. The housing sector plays an important role in the formation of the business cycles. Strong housing demand reflects the condition of the performance of any economy. Increase in the housing demand and housing starts reflects the confidence in the economy. Throughout recent economy, the housing sector surprised many economists with its amazing strength. At the same time home builders continued to meet that demand and as a result the investment in the housing sector never faltered throughout the recession. This performance strayed out of the norm; this prompted many economists to do further work on the subject. The housing boom has led to the increase in employment in the construction sector while other manufacturing and service sectors are facing severe job losses. Moreover, home owners have been able to finance higher priced housing by taking advantage of lower interest rates.

The number of housing starts is influenced by a number of factors some demographic and others economic. Demographic changes like huge migration, increases the housing starts in a particular area. The median age of the population also affects the demand for starts. The baby boomers have been a dominant demographic factor in the last 20 years. The aging of the baby boomers have led many economists to speculate that housing starts will decrease over time and hence housing prices. Ethnicity, family type, and immigration or native born statuses are also some important demographic factors affecting housing starts. An increase in the mortgage rate charged by the banks decreases the activities of housing starts. The interest rate determines the houses being sold. The interest rate is an important variable in determining the decision of a household in buying houses. We will try to see if interest rate affects housing starts significantly. An Increase in the income level of the household helps families to invest in homes. The production index for construction of houses also has its affects in the number of housing starts. With an increase in the production index, the cost of building homes increases, hence decreasing the number of housing starts. I will be reviewing various studies done on the topic to come out with my own model to explain housing starts in America.

Literature Review

Lawrence Smith (1969) in his paper "A model of the Canadian housing and mortgage markets" has used other research findings to develop a complete housing and mortgage model for Canada, with particular emphasis placed on the role of financial variables. He points out that the volume of new house construction actually undertaken depends upon the profitability and the availability of mortgage credit. An increase in User demand for housing unit's means the vacancy rate declines and housing prices rise relative to the construction prices, assuming that the increased building activity does not increase the construction cost in the same period. These variations, Smith points out, are more likely to be profitable and hence be undertaken. Housing starts also depend upon the cost and availability of mortgage loans. In his model he has used housing starts as a function of housing prices, the vacancy rate, construction and the land cost, and the cost and the availability of private and public mortgage credit. The regression model he used has been modified because of the data limitations and institutional consideration. He has deleted the vacancy rate from the model because of the unavailability of the data in Canada. Availability of mortgage credit has been represented by the yield differential between mortgages and bonds, a dummy variable has been introduced to represent the government winter house building incentive program. The housing prices have been represented by the housing prices to construction cost ratio. The mortgage bond yield differential has been used to represent private credit rationing since the supply of mortgage credit from the private financial institutions appear to be quite sensitive to this differential. He has used quarterly data from the 1954-65 periods.

His result shows that the ratio of housing prices to construction to land cost seems to exert a substantial influence on the housing starts, as do the availability of private and public mortgage credit and the cost of this credit. His results show that the "availability of private and public mortgage variable" increase of one million dollar of central mortgage and Housing Corporation lending in 1975 dollars, generates additional eighty seven housing starts. This model has been successful in explaining the housing starts in Canadian market.

Rodney Thom (1985) in his paper "The relationship between housing starts and the mortgage availability" presents a model of housing starts using a housing mortgage availability variable. He has used monthly US data over the years 1967-1984. He points out that the cyclical fluctuations in home building continue to be determined largely by the pattern of deposit flow in the thrift institutions. There are various studies done to examine if mortgage availability affects the number of housing starts. In his model he has used average interest rate on long term government bonds, interest rate on new home mortgages, mortgage availability and number of privately owned housing starts as independent variables. He has defined mortgage availability as the change in the real deposits flow through savings and loans association and mutual savings banks. His result shows that housing starts exhibit significant responses to shocks in the interest rates and mortgage availability. His result shows that responses to the availability shocks are less significant than the response to the mortgage rate shocks. However the result clearly shows that mortgage availability does play an independent and significant role in explaining housing starts. He concludes that mortgage availability is by no means the dominant influence in the housing starts and its role is no greater than that played by the interest rate.

Simone Clemhout