The intra-urban residential and workplace locations of small business owners

ABSTRACT

The notion that buzz, creativity, diversity, openness, and a sense of bohemia in cities are important to attract creative workers and entrepreneurs has grown in prominence both in academic literatures and in city economic development strategies. However, there is a disjuncture in the literature and a dearth of evidence as to whether entrepreneurs seek bohemian (open, diverse) places in which to live or to locate their business. This study explores the kinds of neighborhood small business owners, in particular entrepreneurial small business owners, live and work in, and the extent to which their intra-urban locational patterns diverge from the general working population. Survey data of small business owners in Edinburgh (UK) uniquely capturing both business location and the residential location of the business owner, and census data covering all workers with workplaces in Edinburgh are used. Findings support the attraction of some entrepreneurs to bohemian neighborhoods both as places to live and as places to work. Equally, however, findings stress the importance of a diversity of neighborhood types, including attractive suburban neighborhoods, due to business cycle and personal life course effects, making non-bohemian neighborhoods also attractive to small business owners.

Acknowledgments

We are indebted to Martin Schäfer (University of Portsmouth) for the selection and extraction of bespoke commuting flows from the Census of Population and Paul Carter (University of Portsmouth) for creating the Edinburgh map. We are grateful to three anonymous reviewers for their valuable comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. National Records of Scotland, Scotland’s Census 2011 Table KS204SC.

2. These figures only include registered businesses.

3. ONS Mid-year population estimate.

4. It is estimated that three-quarters of all UK enterprises in 2015 were one-person businesses and that the majority of these were unregistered businesses (56%) and/or home-based (59%) (BIS, 2016).

5. ONS Inter-Departmental Business Register.

6. Businesses were not included if information on the number of employees was missing and when information on either the residential or business location was incomplete or not within the sampling framework.

7. ONS UK Standard Industrial Classification 2007.

8. Eurostat indicators on High-tech industry and Knowledge-intensive services (Annex 8).

9. “Data Zones” in Scotland are similar in spatial scale and definition to “Super Output Areas” in the rest of the UK.

10. Variables that were used to derive neighborhood types are published in Gale et al. (2016). However, information about the precise composition of each neighborhood type is not available.

11. Where appropriate, we follow ONS’s eight “Super Groups” with the following alterations: (i) “Cosmopolitan student neighborhoods” and “Inner-city cosmopolitan” combined to form “Cosmopolitan,” (ii) “Multicultural living” and “Ethnically diverse professionals” combined to form “Ethnically and culturally diverse neighborhoods,” (iii) the following groups are reported as distinct categories rather than being aggregated within their larger “Super Group”: “Highly qualified professionals,” “Affluent communities” (which we rename “Affluent suburban communities” to reflect its Super Group) and “Comfortable suburbia,” and (iv) the “Aging suburbia” Group is added to the “Countryside Living” Super Group and renamed “Aging/rural.”

12. As well as being dominated by “cosmopolitan” neighborhoods, the zone we define as central Edinburgh (within 2 km of the central railway station) represents the city’s high-density core, thus captures an important spatial structure of the city in relation to conventional urban models of residential choice. Population density drops by 35 persons per ha (from 81 to 46) between the 1–2 km and 2–3 km rings, by far the greatest single drop in population density moving out from the city center.

13. This approach also conforms with sample size rules of thumb suggested by Tabachnick and Fidell (2014, p. 159).

14. We define Central Edinburgh as all Census Data Zones with their center falling within a 2 km buffer of the city’s main railway station on Waverly Bridge, which lies at the heart of the central business district. Outer Edinburgh is defined as the area between the central zone and the city’s administrative boundary. Beyond the City of Edinburgh administrative boundary lies our final zone, “outside Edinburgh.” See also endnote 12.

15. The level of in-commuting is of a similar magnitude to the 38% of jobs in the 12 English Local Authority areas with >100,000 workplaces plus Aberdeen and Edinburgh in Scotland, Belfast in Northern Ireland and Cardiff in Wales (own calculation based on 2011 Census of Population).

16. Business owners’ commute distances were calculated as straight-line distances between the centroids of postcode unit polygons. Postcode units contain an average of 15 addresses each and rarely more than 100. For all workers, we calculated the commutes in the 2011 Census of Population as straight-line distances between the centroids of Census Data Zones. From the Census dataset, 97,199 recorded commutes were extracted originating within Scotland (from residences in a total of 6,976 distinct Data Zones) and ending in one of the 115 Census Data Zones in Edinburgh which contained at least one business address from the present business survey (Table WF01BSC_DZ2011_Scotland).

17. Commuting distances to workplaces in Edinburgh is in line with comparable cities. The average commute distance of workers with workplaces in Edinburgh is 11.6 km. The average commute to workplaces in similar-sized cities in England plus Aberdeen and Edinburgh in Scotland, Belfast in Northern Ireland and Cardiff in Wales is 12.5 km (own calculation based on Census of Population 2011 Table WP702EW and Table LC7102SC from the Scottish Census).

18. Probability = elogit/(1+ elogit).

19. Standard Industrial Classification (SIC) codes K, L and M combined.

Additional information

Funding

The data collection for this research was funded by an Early Career Grant from the Regional Studies Association to Darja Reuschke. Darja Reuschke’s time working on this paper was funded by the WORKANDHOME Starting Grant from the European Research Council [ERC-StG-2014-639403].

Notes on contributors

Darja Reuschke

Darja Reuschke is Associate Professor in the School of Geography and Environmental Science at the University of Southampton. Her research focuses on home-based businesses, self-employment and micro enterprises and how these impact on cities and are influenced by various economic and social processes such as digitalization and societal modernization.

Donald Houston

Donald Houston is Professor of Economic Geography at the University of Portsmouth. His main research interests relate to how urban and regional change affects employment outcomes, including the mediating effects of commuting.