ABSTRACT
The construction sector in Turkey boomed in the period between the turn of the millennium and the onset of the global crisis. This paper studies the employment generation effects of growth of this sector, taking into account that it not only depends on domestic production factors but also on imported inputs. In order to unravel the interactions of the construction sector with the rest of the economy, structural path analysis of both labour and imported intermediate input demand generation is used based on the 2002 and 2009 WIOD data for Turkey. The findings indicate that labour linkages weakened between 2002 and 2009, while import linkages became slightly stronger. The sectors that have played key roles in this are identified, as are the linkages between these. The findings have implications for the persistent unemployment and current account deficits in Turkey.
Acknowledgements
We thank Cemre Senesen, Tunca Ulubilge and Murat E. Unal for their help at the early stages of software development. Comments of the editor Bart Los and the two anonymous referees towards improvement of the text are deeply acknowledged.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 See, e.g. Crowe et al. (Citation2011, p. 42, 50–57) for an overview.
2 At the time, TOKI also started the planning for a second batch of 500,000 dwellings after 2010 to be undertaken by large-scale private companies.
3 Turkish employment data for 2009 were the most recent data available when we did the calculations.
4 The journal Construction Management and Economics published articles using input–output analysis of the construction sector which involve simple linkage multipliers for various countries. See also Kronenberg et al. (Citation2012), who find that construction-related activities initiated by the government, like the refurbishment of existing buildings for the case of Germany, have significant employment effects. This study suggests that this policy might also help troubled economies such as Spain and Ireland in crisis management.
5 See Senesen et al. (Citation2011) for a survey of implications of jobless growth in the context of interindustry transactions.
6 The source of data in this section is TURKSTAT: www.tuik.gov.tr, www.turkstat.gov.tr
7 This share (computed as value added in construction divided by GDP) is higher than the value for the USA just before the crisis in 2007 (5%, see Nolan, Citation2011) but lower than the value for Spain in 2008 (10.7%, see Garriga, Citation2010; Perez, Citation2010).
8 This data set is available since 2000. See www.turkstat.gov.tr.
9 The latest available I–O table by TURKSTAT at the time of this analysis was for 2002. We use the WIOD tables for data consistency in time. SPA findings with TURKSTAT 2002 I–O data in an earlier version of this paper are qualitatively similar to the findings with WIOD data. See Gunluk-Senesen and Senesen (Citation2011) for a discussion of estimation problems associated with the sectoral employment data for 2002.
10 The labour threshold values in the SPA analysis are 73.861 for 2002 and 35.908 for 2009.
11 The CON → AGR path is rather extraordinary in one respect: There is no direct route between the two sectors. As Figure clearly displays, two different routes, CON → TRD → AGR and CON → WOD → AGR, connect CON and AGR. The impact of investment in CON on employment AGR is not direct but through the influence of CON on TRD and on WOD.
12 The intermediate import input threshold values are 160.671 for 2002 and 167.974 for 2009.
13 There might be a slight decrease in real $ terms. A partially offsetting factor would be a higher overvalued Turkish Lira in 2009, see the discussion in Section 6.1.