Legal environment and new firm formation: case of Turkey

ABSTRACT

The new Turkish Commercial Code decreased transaction costs for joint stock corporations relative to limited liability companies, and led to a marked increase in the number of new firms adopting the corporation organizational form. We examine whether there has been a structural change in relative frequencies of corporations versus limited liability companies due to the new commercial code of 2012. We employ Bai and Perron’s multiple structural breaks test for the period 1995:01 to 2019:01. We find two breaks when we consider only the mean process, and one break that corresponds exactly with the new Turkish Commercial Code when we consider the mean and autoregressive processes. We conclude that the new commercial code increased the number of new joint stock corporations relative to new limited liability companies.

KEYWORDS:

• Firm demography
• structural break
• firm entry regulation
• institutional change and entrepreneurship

JEL CLASSIFICATION:

• D21
• D23

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ There are simpler procedures in linear models when there is a single break and the break date is known a priori (Quandt 1960) or is unknown (Andrews 1993; Andrews and Ploberger 1994). When the underlying model has nonlinearities then smooth transition and threshold models can be estimated (Dijk et al. 2002). Lasso (Tibshirani 1996) models simultaneously estimate the number of breaks and break points in both linear and non-linear contexts but these models to date have not been successfully applied in settings that do not involve i.i.d. errors (Casini and Perron 2018)..

² https://www.tobb.org.tr/Sayfalar/Eng/AnaSayfa.php.

³ We have also considered the difference between new corporations and new limited liability companies but we strongly prefer the ratio as it is a scale-free measure whereas the difference is a measure that will get larger as the economy grows over time. With this measure the inception of the new Turkish Commercial Code is again one of the break points, however, this measure produces more breaks associated with major economic and political events in Turkey. The results are not presented here but are available upon request..

⁴ The period under study is also a period of structural change which is reflected in sectoral firm formation rates. For example, the proportion of new firms has been increasing in trade and declining in manufacturing. The construction sector was one of the drivers of the Turkish economy up until the recent downturn in the real estate market. However, sectoral data are only available starting in 2010 so we cannot undertake a sectoral analysis for the period under study. Preliminary analysis on sectoral firm formation rates for the period after 2010 is not indicative of sectoral variation in the ratio of new corporations to LLCs..

⁵ The choice of the maximum number of breaks is determined by the trimming value we have chosen. With a trimming value of 0.2, the maximum number of breaks we can test for is three.