Move histories and socio-economic position

Neighborhoods are hot in the Netherlands. Especially the problems that have arisen in some neighborhood have attracted a lot of governmental attention the last couple of years. Reason for me dive into some of the literature on residential segregation, troubled neighborhoods, and obviously the people living in these neighborhoods.

I have argued elsewhere that to understand the state that neighborhoods are in on account of whatever characteristic, it is crucial to focus on the individual residential mobility histories of the inhabitants of the neighborhood. The paper discussed today does exactly that by attempting to answer two research questions: (1) “what are the main differences between the migration histories and dwelling careers of different socio-economic groups?” and (2) “what is the relation between dwelling careers and urban structure?”

The conclusions drawn in this paper by Harts and Hingstman may at present seem a little straightforward, but remember that it is a 1986 paper. The conclusions are clear and add some to our understanding of differential residential mobility patterns in the Netherlands, especially because it has taken up a life-course perspective.

The first conclusion the authors draw is that there are indeed differences to be found between mobility patterns of the three SES groups. Low, middle, and high SES groups are distinguished. However, despite high levels of social mobility in the Netherlands during the period investigated, the great majority of
households climb only a few steps on the housing ladder. Nevertheless, the higher the social economic status of a person or family, the better the starting point regarding the residential mobility career.

Regarding the second research question it is argued that the Dutch legislation regarding housing determines the options people have when selecting a place to live. Especially low-priced housing is managed by the government, often determining that only people with economic ties to a specific city are allowed to live in the low income houses. Thereby, “the characteristics of the housing market and urban structure influence the course of the dwelling careers, the characteristics of the dwellings, but also the spatial path. It is a continuous matching process of dwellings and population”. Unfortunately, the effects of this are not shown directly; only a possible consequence is shown: people tend to move relatively often within the own city. It is not clear whether this especially holds for people of a low SES, which would make a very interesting hypothesis.

There are, however, some drawbacks regarding the method of investigation employed in this paper. For starters, the analyses are based on a sample of Dutch couples, living in the Dutch city of Utrecht at the moment of data collection (1982). Couples apparently needed to be married, and have done so during the 1960’s pr 1970’s. Data was collected by means of retrospective questions. Thereby, I do not feel that the conclusions could be inferred to the complete population of Utrecht, for clearly not only married couples lived in Utrecht. Also, no attention is paid to the effect of possible selection effects, due to migration into the city or due to divorce or death.

A more serious issue I have with this paper is that several statements are made that are not backed up with any empirics. Presumably, this paper functions as a summary of a PHD project, but nevertheless I do not trust any of these statements without seeing some tables. Unfortunately, this addresses some of the most interesting statements made in the publication. For instance, it is argued that the number of moves during the life-course of a couple averages between five and six, but this is not shown. Furthermore, it is subsequently argued that this average is primarily differentiated by social economic class. However, the impact of determinants other than SES is not shown.

The method of analyses is primarily bivariate, which is unfortunate because many of the important covariates in a life-course perspective (such as year, or age) are not taken into account, nor are variables indicating the position in the life-course a couple is in, such as for instance the number of children the couple has. Combined with the categorization of social economic status in three groups, it becomes difficult to distinguish the actual causal mechanism at work. For instance, in the text it is argued that income is an important determinant of the quality or type of housing a person lives in. Although I’m willing to believe this at face value, it is not empirically tested in the tables, for they only show the impact of social economic status, which encompasses more than mere income.

Despite the drawbacks mentioned, I still think that this paper has contributed to our understanding of residential mobility patterns. The weakest point of this study is probably the operationalization and method of analysis, due to which no causal inferences could be made and due to which many of the more interesting statements made were not tested empirically. To the contrary, the life-course perspective was, I think, even a little ahead of its’ time in 1986. Also, the authors have managed to give a nice overview of theoretical thought on the subject, although not all are empirically tested. But most of all, I find this a strong starting point for further investigation of differentiated residential mobility patterns in the Netherlands.

Harts, J.J., Hinstman, L. (1986). Move histories and socio-economic position. Journal of Housing and the Built Environment, 1(4), 343-352.

Leave a Reply