How many jobs does construction expenditure generate?

Michael Ball, Andrew Wood

Research output: Contribution to journalArticlepeer-review

20 Scopus citations


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.

Original languageEnglish
Pages (from-to)307-318
Number of pages12
JournalConstruction Management and Economics
Issue number4
StatePublished - Jul 1 1995

Bibliographical note

Funding Information:
We should like to thank the Joseph Rowntree Foundation and the Department of the Environment (DOE) for financial support; staff at the DOE, Julian Aubert, Alison Booth, Ron Smith and the anonymous referees for assistance andor comments. The opinions expressed and any errors remaining are the responsibility of the authors alone.


  • Employment
  • costs
  • public expenditure

ASJC Scopus subject areas

  • Management Information Systems
  • Building and Construction
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'How many jobs does construction expenditure generate?'. Together they form a unique fingerprint.

Cite this