Institutions, financial development, and small business survival: evidence from European emerging markets

https://​ec.​europa.​eu/​growth/​smes_​en. A similar pattern can be observed in the USA, where 28 million small businesses account for 99.7% of all firms. Moreover, family firms, which are often SMEs, constitute the pillars of the economy in most countries (La Porta et al. 1999).

SMEs were among those firms facing very difficult challenges, and large numbers had to exit the market—the uncertainty regarding the economic policies and regulations in which enterprises operate has increased dramatically since the outbreak of the GFC. As a result, SMEs were probably the hardest hit economic units due to their limited capacity to absorb large shocks as well as their sensitivity to uncertainty in the economic environment (Potter and Thompson 2011).

The European Commission (2003) defined small- and medium-sized enterprises (SMEs) as companies with fewer than 250 employees that operate independently from larger companies. To be classified as an SME, a company should have a maximum annual turnover of €50 million or a maximum annual balance sheet of €43 million. There are three types of SMEs: micro-SMEs, with fewer than 10 employees; small SMEs, with 10–49 employees; and medium-sized SMEs, with 50–249 employees (https://​ec.​europa.​eu/​growth/​smes/​business-friendly-environment/​sme-definition_​en).

The Orbis database keeps all entry firms with their legal status—including bankrupted companies—without any time limitation unless local data providers stop supplying company information. We carefully checked the original source of our data and detected that as of 2017 (the end of our research period), data are available for more than 95% of firms registered in the 2006 archive. Hence, only a very limited number of firms were dropped from the Orbis database during the observation period of 2007–2017.

In order to address concerns regarding reverse causality, authors typically estimate their models using financial development indicators in the initial year of the estimation period. This is exactly the approach we take, and we employ 2006 values as predetermined variables both for the country-level and firm-level covariates to avoid any endogeneity issues. This is also a methodological standard of the survival analysis based on the Cox hazards model. An alternative approach in the literature is to use annual values of financial development indicators to avoid significant loss of information. We do not adopt this alternative approach, as it does not fit into the methodology we use for our analysis, since the basic aim of survival analysis is to test whether an initial condition is a good predictor of the event in question.

In our paper, we accentuate the role of two “macro” dimensions, namely, institutional quality and financial development. In order to deal with the omitted variable bias, we also considered other macro-factors, namely, GDP growth rate, CPI annual average, GDP per capita, population size, and current account to GDP. All factors produced statistically significant hazard ratios, with the exception of the current account to GDP (not reported). These additional macro-variables impact the probability of firm survival within a range of 1–9%. Hence, their economic significance is decisively smaller than that of individual characteristics of institutional quality and financial development reported in Tables 3 and 4. The use of a 3-year average (2004–2006) generated similar results. As a result, we focus on the effect of IQ and FD variables to assess the survival of SMEs in European emerging economies.

We also ran regressions introducing the square of the firm size but obtained statistically insignificant results. It follows that firm size does not seem to exhibit a nonlinear relationship with respect to small business survival probability in European emerging markets.