THE GEOGRAPHY OF FIRM DYNAMICS -

MEASURING BUSINESS DEMOGRAPHY

FOR REGIONAL DEVELOPMENT

OECD Working Party on Territorial Indicators Paris, 5 December 2017

• Assessment of the main methods in measuring

subnational business demography

• Development of a new database for OECD

regions

• Analysis of key issues and trends in business

dynamics at the regional level

Objectives of the Project

• New businesses constitute, on average, 10% of all firms but regional variation is substantial (~9% for employer firms, 12 countries)

– Concentrated in urban and most productive regions (frontier)

– Associated with better local governance, more developed R&D infrastructure, lower financing constraints, and more educated local labour force

• Exact information on location of employment is important for understanding the impact of business dynamics on employment

– Using enterprise-level data can be susceptible to a bias from a region’s actual share of national employment

– Headquarter bias: pronounced in capital-city regions

• New businesses contribute significantly to regional employment growth

– Business births can create up to 8% new employment

– Higher regional firm dynamics also indirectly boost employment growth in existing firms

Key Findings

Business births, TL3 (2014) Capital regions drive business creation

4

• On average, 10.2% new firm but large cross-country and within-

country variation.

• Firms births slightly lower for employer firms at 9.4%.

Employer vs. non-employer firms Large country differences in importance of non-employer

firms

• Different national tax codes and legal systems

• Ratio of regional birth rates in employer and non-employer firms not

always strong

• Most consistent and comparable approach: employer firms only

Share of employer firms by country

Deaths and births, TL3 Highly urbanized regions are more dynamic (higher churn)

6

Birth and death rates as well as shares of births, deaths and active firms

by degree of urbanization, all firms

Urban regions more dynamic (larger birth and death rates) and

accounting for disproportionate share of firm deaths and especially

births

Breakdown by sector

Urban-rural difference reflects sectoral composition of regional

economies

7

Birth shares by sector (TL3, 2014, or last available)

• Large regional variation in employment creation rates, ranging from

8% to 0.2%. Even within countries, differences are considerable.

• Great potential for job creation, enhanced by indirect effects

Employment creation by new firms, TL3 Significant within-country differences in created employment

Employment creation rates by new employer enterprises (2014)

Indirect employment effects of firm dynamics Employment growth in existing firms

• Interaction of firm dynamics (churn of firm) and EU structural

funds:

• Firms that received EU structural funds grew more if they were

located in regions with more dynamic firm environments

• Indicative of indirect effects – higher firm dynamics increase

competition & efficiency

The headquarter bias: differences in national

employment shares (2014, TL2) Illustration with capital-city regions

• Enterprise-level data causes a 7 percentage point upward bias in

employment statistics of capital regions (~1.4 p.p. on average)

𝐻𝑄_𝐵𝑖𝑎𝑠𝑖𝑐 = 𝐸𝑚𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑖𝑛 𝐸𝑛𝑡𝑒𝑟𝑝𝑟𝑖𝑠𝑒𝑠𝑖𝑐𝐸𝑚𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑖𝑛 𝐸𝑛𝑡𝑒𝑟𝑝𝑟𝑖𝑠𝑒𝑠𝑐

−𝐸𝑚𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑖𝑛 𝐸𝑠𝑡𝑎𝑏𝑙𝑖𝑠ℎ𝑚𝑒𝑛𝑡𝑠𝑖𝑐𝐸𝑚𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑖𝑛 𝐸𝑠𝑡𝑎𝑏𝑙𝑖𝑠ℎ𝑚𝑒𝑛𝑡𝑠𝑐

Which regions are more dynamic in terms

of business creation/destruction?

11

• Regional drivers of firms’ dynamics:

What institutional/social factors are driving business

creation/death/survival ?

• Regional business structure: dominance of a few large firms

• Institutions (corruption, business regulation)

• Credit constraints / funds

• Human capital and R&D (innovation)

Dominance of large firms and birth rates -TL3 Different development models in urban and rural regions

• Entrepreneurship in urban and rural regions is differently linked to the

local business environment and cluster formation

• In urban areas, clusters dominated by a few champions thrive the most

• Average birth rates between 2008 and 2015. 165 TL2 Regions of 15

countries.

• Quality of local governance is positively correlated with firm

creations

Local governance and firm birth rates Gallup World Poll Indicators (average 2008-2015)

• Large heterogeneity in business demography across regions – Discrepancies between urban and rural regions

– Differences along the lines of regional productivity

• (At least partially) explicable by regional characteristics: – Local governance, credit constraints, education and innovation,

concentration of firms

• Monitoring regional employment creation by new firms requires precise geographic information – Best available data show that the contribution of new enterprises is

large and very heterogeneous across regions

– Headquarter bias can lead to deviations from actual employment creation

Conclusions

Statistical agenda: - Further harmonise the production of regional business

demography indicators, building on the enterprise approach and possibly distinguishing employer firms and at the sufficiently detailed geographical level (TL3)

- Submit an update for the OECD-Eurostat Entrepreneurship Indicator Programme for the measurement of regional business demography indicators

Analytical agenda: - Using large scale firm level data set (i.e. ORBIS) to estimate the role of local characteristics associated with the success of firms - Further focus on the performance of SMEs and their role for regional development

Further steps

THANK YOU Contact: paolo.veneri@oecd.org lukas.kleine-rueschkamp@oecd.org