« Previous | Contents | Next »
Listen
1. INTRODUCTION
In the summer of 2007, the Scottish Government published market evidence for Scotland's housing market known as the Scottish Housing Market Review (Scottish Government, 2007a). This analysis indicated evidence of increasing supply inelasticity (a lack of response by new supply to rising house prices). It also suggested a build up of affordability problems in different local markets across Scotland. The evidence presented, while more suggestive than that thoroughly researched from a comprehensive modelling process, also indicated that housing supply would have to increase substantially (from around 20,000 units per annum to 35,000 units) in order to have affordability ameliorating consequences. These figures were derived from an extrapolation to Scotland of the North East and North West England regional affordability models run for the Department for Communities and Local Government ( DCLG). The figures subsequently formed the basis for the policy aspiration set out in Firm Foundations (Scottish Government, 2007b) to seek to achieve such an increase in long term average annual completions by the middle of the next decade. However, the consultation document recognised that this house building aspiration had to be thoroughly evidenced by new research that modelled the long term housing market in Scotland (and local sub-regional markets thereof) and allowed for different policy scenarios and exogenous shocks to the systems. This report details the development, estimation, programming and outcomes of that model building exercise.
The Government's Review recognised that there is a lack of coherent economic modelling of the housing market in Scotland to either parameterise the local (i.e. sub-regional) and national housing markets or to clarify affordability and housing supply relationships in the long term and across different plausible policy scenarios or shocks to the economy or housing system. The Government decided what was needed was no less than a fully-functioning simulation model of the Scottish housing market, in the spirit of the DCLG regional affordability simulation model (Meen, et al, 2005), and that operated at functional sub-Scotland geographies. It was recognised that this was a large undertaking in terms of research over a relatively short period of time (January to September 2008). It would require several major distinct tasks that would have to be subsequently integrated: development of a conceptual model, development of a series of interacting modules (household formation, migration, tenure choice, labour market and house price determination), agreement on sub-Scotland geography, data collection and estimation, programming and testing of the full simulation model.
The research team, led by Dr Chris Leishman, successfully completed the research task in terms of these requirements. However, the team is clear that this is only the first stage of a process of model building, refinement and revision that needs to go forward. By its nature, the model presents a long term picture of the Scottish housing market that abstracts from short run deviations. While it is perfectly possible to build a scenario based, for instance, on expected duration of a recession or the credit crunch, even building in official economic forecasts and to proxy its effects, these remain highly schematic ways of developing such scenarios. Long term modelling plays a certain role in helping develop policy but it is always at the mercy of rapidly changing economic circumstances such as those visiting the global economy at present. Other voices criticise the assumptions underpinning these complex simulation models and the model builders need to respond by improving estimating equations, enhancing the transparency of the model and working with the best and most up to date data. The English regional model has been in a near constant state of updating and modification since it was first developed. The Scottish model has learned much from its English counterpart but more than anything the research team have recognised the importance of continually seeking to improve the model.
This report provides a comprehensive overview of the model and its application. It starts in section 2 with a summary of the modelling approach undertaken and a short overview of the strengths and weaknesses of the modelling strategy. Section 3 details the modelling theoretical framework and how the sub-national geography was arrived at. Section 4 sets out the results of the model for each module and the combined simulation model. Section 5 reflects on possible future modelling extensions that will strengthen the model's usefulness. Section 6 concludes.
« Previous | Contents | Next »