Jacob Veenstra

Scale, mergers and efficiency: the case of Dutch housing corporations

The efficiency of social housing providers is a contentious issue. In the Netherlands, there is a widespread belief that housing corporations have substantial potential for efficiency improvements. A related question is whether scale influences efficiency, since recent decades have shown a trend of mergers among corporations. This paper offers a framework to assess the effects of scale and mergers on the efficiency of Dutch housing corporations by using both a data envelopment analysis and a stochastic frontier analysis, using panel data for 2001–2012. The results indicate that most housing corporations operate under diseconomies of scale, implying that merging would be undesirable in most cases. However, merging may have beneficial effects on pure technical efficiency as it forces organizations to reconsider existing practices. A data envelopment analysis indeed confirms this hypothesis, but these results cannot be replicated by a stochastic frontier analysis, meaning that the evidence for this effect is not robust.

Relative efficiency, mergers and the scale paradox: the case of Dutch housing corporations

The efficiency of social housing providers is a contentious issue. In the Netherlands, there is a widespread belief that housing corporations have substantial potential for efficiency improvements. A related question is whether scale influences efficiency, since recent decades have shown a trend of mergers among corporations. This paper offers a framework to assess the effects of scale and mergers on the efficiency of Dutch housing corporations by using both a data envelopment analysis and a stochastic frontier analysis, using panel data for 2001–2012. The results indicate that most housing corporations operate under diseconomies of scale, implying that merging would be undesirable in most cases. However, merging may have beneficial effects on pure technical efficiency as it forces organizations to reconsider existing practices. A data envelopment analysis indeed confirms this hypothesis, but these results cannot be replicated by a stochastic frontier analysis, meaning that the evidence for this effect is not robust.