How many jobs does construction expenditure generate?

Extra public expenditure on construction is a frequently cited policy to alleviate demand deficient unemployment; yet the actual number of jobs created is uncertain. This paper surveys the job creation effect for the UK. It surveys previous estimates, then a construction employment model is estimated. The results do not conform to theoretical predictions, implying only a weak link between quarterly increases in total construction output and construction employment, and no significant relationship between housebuilding and employment. It is suggested that these results arise from poor quality data, especially the estimates of changes in the number of self-employed workers. As an alternative, construction industry rules-of-thumb are used to derive more plausible employment effects. Employment estimates are also provided for subsectors of the industry. To improve on these calculations we conclude that there is a need for either more accurate construction data or, failing that, periodic site-survey based estimates of construction production functions.