Growth, stability or decline in rural enterprises - Michael Ridge, Director, Frontier Economics
1. Drivers of rural business growth, decline
and stability
CONFIDENTIAL
This work contains statistical data from ONS which is Crown Copyright. The use of the ONS statistical data in
this work does not imply the endorsement of the ONS in relation to the interpretation or analysis of the statistical
data. This work uses research datasets which may not exactly reproduce National Statistics aggregates
2. Rural and urban definitions used in the
analysis
Rural-urban classification Sparsity classification
Classification
at OA level
3. Rural and Urban are not very different…
0%
5%
10%
15%
20%
25%
30%
Agriculture/fishing
Mining/quarrying/
construction/utility
Manufacturing
Wholesale/retail/
logistics
Services
Finance
Health/socialwork/
education/public
Tourism-related
Shareofemployment
Rural Urban Less sparse Sparse
4. …but Rural performed better
95
96
97
98
99
100
101
102
103
2008 2009 2010 2011 2012
Employmentinarea(2008=100)
Rural
Urban
Less
sparse
Sparse
Total
5. Which sectors drove the difference?
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
Rural Urban Less sparse Sparse
Contributiontochangeinemployment(2008-2012)
Tourism
Health/ social work /
education / public
Finance
Services
Wholesale/ retail / logistics
Manufacturing
Mining/ quarrying/
construction / utility
Agriculture/ fishing
Overall net change
6. Top 10 growth sectors Top 10 decline sectors
Social work without accommodation Employment activities
Residential care activities Construction of buildings
Education Civil engineering
Retail Manufacture non-metal mineral
Human health activities Manufacture fabricated metal
Agriculture Manufacture motor vehicle
Service to buildings Office admin, business support
Computer prog, consultancy Manufacture plastic / rubber
Sports & recreation Repair installation machinery
Telecommunications Manufacture machinery nec
Which sectors have changed the most in
rural areas?
These are a mix of cyclical sectors (e.g.
construction / business support) and
manufacturing
Main growth is in social work and
residential – both key sources of
employment growth. Other growth
sectors are more of a mixed group
7. Rural performed better in each region…
-6%
-4%
-2%
0%
2%
4%
6%
NorthEast
NorthWest
Yorkshire
EastMidlands
WestMidlands
EastofEngland
SouthEast
SouthWest
Changeinemploytment(2008-2012)
Rural Urban
8. Why do rural businesses do better in some
places than others?
Supply side Demand side
● Input costs (diverse and
vary by industry)
● Labour: price + quality
● Transport / transactions
costs
● Local demand drivers
(population, income)
● Search goods (customers‟
knowledge of product)
● Wider e.g. global demand –
exchange rates etc.
9. But for reliable analysis you need data at a
fine spatial scale
Postcode / OA level data
c. 170k OAs
LAD level data
Only 326 LADs
● Broadband
● Population, households and
income
● Transport infrastructure
● House prices
● Demographics – pensioners
and graduates
● Number of public sector
employees
● Measures of sectoral
agglomeration
● Business rates
10. Quantitative analysis found some drivers
related to rural business performance…
Broadband / mobile
infrastructure
● Rural areas with good broadband have fared a lot better
● Strongly correlated with business births and employment
growth
● Also linked to churn…
Local demand
● Population, households and income are all correlated with
business performance
● Effect seen in relation to both changes and levels; the latter
effect could suggest some form of agglomeration effects
occurring (relatively more growth in more populated areas)
Transport
infrastructure
● Better connected areas have higher rates of both start-ups
and failure (i.e. higher churn)
● More flexibility to re-deploy inputs in other uses, so this
should enhance productivity
11. Qualitative analysis confirmed many of these
findings and identified other areas of concern
Broadband/
MobilePhone
Coverage
Transport
Infrastructure
AccesstoFinance
● Access and speed
□ Slow broadband constitutes a major time cost,
both waiting for up/downloads and/or traveling
to use the internet
□ High cost of installing lines in rural areas
□ Barrier on tourism as free Wifi expected
● Roads
□ Poor public transport options perceived as a
major barrier to employing low level staff as
they cannot afford to run a car
□ Closures can be costly in terms of customer
access and time cost for employees
● Rail
□ Seen as a driver in North-East/North-West and
a major barrier in South-West
● Participants felt investment capital was unavailable
□ Banks requiring too much security on loans,
loss of the „local‟ bank manager
□ Most had written off commercial finance and
were extremely adverse to debt. They were
therefore expanding from profits alone
Loansand
Grants
Planning
Employmentand
Employmentlaw
● Lack of awareness
□ Grant application process too complex with multiple
organisations. Participants were put off by this
□ No real marketing of peer-to-peer lending schemes
● “Can‟t do”, “Detached from reality”
□ Participants felt that planning authorities were not
on their side and as a result there was too much of
a time burden on senior staff
□ Planning authorities unrealistic about stopping
development on green land
Employees need cheep housing
● Payroll and Auto-Enrolment
□ Frustration at the time cost on senior management
at having employees on payroll and auto-enrolment
● Access to Labour
□ There was a wide perception that employers weren‟t
getting value for money and employment law was
making it hard to remove poor performers
12. Summary
● Rural and urban overlap considerably in terms of sector. Around 80%
of rural employment is in sectors that are equally important in the
urban economy.
● Rural has performed better than urban (at least as far as
employment), in most sectors and regions.
● Rural growth has been stronger in areas that are in close proximity to
urban areas, suggesting there may be agglomeration benefits in linking
resource-rich rural areas with centres of economic activity.
● Much of the change in rural economic activity can be characterised by
sector-specific structural factors or cyclical factors.
● There is a strong relationships between rural performance and
broadband provision.
● The planning system is claimed to be an over-zealous constraint on
economic development and on housing.
13. Frontier Economics Limited in Europe is a member of the Frontier Economics network, which consists of separate companies
based in Europe (Brussels, Cologne, London and Madrid) and Australia (Melbourne & Sydney). The companies are
independently owned, and legal commitments entered into by any one company do not impose any obligations on other
companies in the network. All views expressed in this document are the views of Frontier Economics Limited.
14. FRONTIER ECONOMICS EUROPE LTD.
BRUSSELS | COLOGNE | LONDON | MADRID
Frontier Economics Ltd, 71 High Holborn, London, WC1V 6DA
Tel. +44 (0)20 7031 7000 Fax. +44 (0)20 7031 7001 www.frontier-economics.com
Editor's Notes
We use the 2001 rural-urban classification, which assigns a rural or urban classification to each Census Output Area (COA) using the 2001 Census. [At the time of this research the RUC11 was not available]. This is a settlement-based approach; an urban COA is one that lies within a settlement with population over 10,000. Rural areas consist of settlements below 10,000 people. Rural areas account for around 18% of the population of England. As well as settlement form, the wider context of each settlement can be determined by looking at dwelling densities to identify sparsely populated areas, and whether a settlement type is sparse or less sparse based on the number of households in the surrounding area up to a distance of 30 km. On different measures, rural areas account for between 13% and 26% of economic activity in England, whilst sparse areas (which will include some rural and some urban) account for between 0.7% and 2.2% of activity.
Here we compare the employment composition by broad sector for different geographies. Each bar shows the share of employment that a sector accounts for in an area type. For example, agriculture / fishing accounts for 7% of but very little urban employment. It accounts for 1.5% employment in less sparse areas - this will come from the rural areas that lie within the less sparse areas.In terms of sector composition, rural and urban are fairly similar: The main sector groups in all area types are in wholesale, retail, services, health, social work, education, and public administration. Nevertheless there are some differences. Agriculture accounts for around 7% of rural employment, but very little urban employment. Tourism-related sectors account for 12% of rural employment, but only 7% of urban. Physically intensive sectors such as manufacturing, construction mining, and quarrying account for more employment in rural. A lower share of rural employment is in services, finance, or various public-related sectors. “Tourism-related” sectors have been defined as: Travel agency, tour operator and other reservation service and related activities; Libraries, archives, museums and other cultural activities; Sports activities and amusement and recreation activities; Accommodation; Food and beverage service activities.
From 2009 to 2011, urban areas experienced a two percentage point decline and then recovered by a percentage point. The trajectories for national average and less sparse are very similar, as they are predominantly urban. By contrast, rural employment levels were stable over that period, seeing only a very modest small decline, followed by a sharp increase of two percentage points from 2011 to 2012. [Assume we don’t want the turnover / productivity discussion]There are a number of hypotheses as to why rural may have performed better than urban:• Lower substitutability. Rural areas will tend to face higher transport costs. This will reduce the range of alternative uses of productive assets. This means there is less to gain from closing or scaling back a business. • Technological change. Over time certain communications and transport infrastructure will have been rolled out to rural areas. This lowers the transaction cost of doing business in rural areas, and may have enabled some areas to unlock growth potential (catch-up or convergence with urban areas). • Structure of employment. Rural employment tends to have relatively more self-employed and home-workers in the labour force. These may add resilience. • Sectoral mix. Some sectors are more rural focused, and some have performed better than others. This could explain some of the difference in rural and urban performance. However, even at individual sector level, rural tends to have outperformed urban, so this cannot be the sole explanation. • Geographic focus of recessionary shock. It is possible that the fundamental recessionary shocks may have been more urban-focused and had less of an impact in rural areas. • More flexible use of assets. It is observed that rural businesses are more likely to be engaged in multiple business activities, so may have more scope to switch between them if needed. For example, farmers might find alternative uses for agricultural buildings if the farming activity is performing less well.
The changes between 2008 and 2012 can be broken down by sector. Each of the rectangles represents the sector group’s contribution to the total change in employment from 2008 to 2012. For example, manufacturing contributed 1 percentage point of decline in rural areas, whilst services contributed 0.5 percentage points of growth. The contributions add up to give an overall net change of + 2% in rural areas (the circles.Some sectors perform similarly in the different area types:Health/social work/education/public was the main contributor to growth, contributing 2 percentage points in rural and 1.5 percentage points in urban. Manufacturing and mining/quarrying/construction/utilities are the main sectors contributing to decline, giving similar reductions in employment in rural and urban areas.Other sector groups perform quite differently in the different area types, so can be thought of as explaining the variation in performance:Wholesale/retail/logistics contributed 0.3 percentage points growth in rural areas, but to 0.8 percentage points decline in urban areas. Agriculture, services and finance each contributed between 0.2 and 0.5 percentage points of growth in rural areas, but had little change in urban areas. Tourism contributed 0.5 percentage points growth in rural areas, but only 0.2 points in urban areas.
We now look at the key growth and decline sectors at a finer level – SIC 2-digit. At this level there are 88 different codes. The 10 growth sectors represent 34% of private rural employment. Between 2008 and 2012 they added these sectors added 108,000 private jobs, representing a 3.4% growth contribution. Many of these sectors were growing both before and during the recession – social work, residential care, retail, services to buildings, computer programming and consultancy, sports and recreation and telecommunications. The 10 decline sectors account for 12% of rural private employment. Combined, these sectors lost 76,700 jobs since 2008, a growth contribution of minus 2.4%. The sectors have a mix of both cyclical patterns and structural change: Employment activities*, construction, civil engineering and office admin had all been growing strongly prior to the recession, but have since been in decline. We could consider these patterns to be cyclical.A number of manufacturing sectors were in decline both in the 2002-2007 period and in the 2008-2012 period – non-metal minerals, fabricated metals, plastic and rubber, and machinery not elsewhere classified. We could consider the decline in these sectors to be structural.StabilityWe also considered what kinds of business were stable over the recession, considering stability in terms of both the extent of entry and exit that occurs (“churn”), and the volatility of employment over time. At a sectoral level, we found land transport and motor trade (wholesale + retail) to be the most stable. We also found churn rates to be lower when further away from major cities, or when a location is less accessible. *The employment activities sector includes both recruitment agencies and temporary employment agencies. We suspect that some of the decline in employment for temporary employment agencies may be due to the EU Agency Workers’ Directive 2008, which may have encouraged employees of the agency to become self-employed.
The regions of England perform differently in terms of rural performance, and also in terms of which sectors perform well. Here we show rural and urban employment growth by region from 2008 to 2012. In every region, rural has had stronger employment growth than urban. The largest difference in performance is in the South West, where rural grew by 5%, while urban declined by 2%. However, in 3 of the regions rural employment did decline –NW, SE, and particularly NE.
We then sought to establish what the key drivers for growth, decline and stability have been since the recession began. At a conceptual level, we can distinguish between factors that affect the level of demand a firm faces, and factors that affect its ability and cost it faces supplying goods and services. The sorts of factors that may be relevant are shown here. The relative importance will also vary greatly by sector. For example, availability of highly skilled staff will be crucial in some sectors but not for others. The question is which of these factors have affected rural business performance in the last few years.
In order to test the impact of these variables empirically, we need there to be local variation. For example, the level of taxes will greatly affect the viability of a business, but unless we can compare businesses facing high taxes with similar businesses facing low taxes, we won’t be able to explore what that effect is. So, which drivers can be examined empirically depends on the data available. Some of the data is available at postcode or OA level. This is very fine and gives a very large effective sample size. Other data are only available at Local Authority District level. There are only around 200 of these with rural businesses, and it is much harder to do obtain reliable results from this.
These are the relationships that we have found, and which remain robust when various changes in methodology are employed. These patterns are empirically robust, but we cannot conclude that they are necessarily causal. We don’t really see the same relationships when we repeat this analysis, but for urban businesses. This would suggest that these sort of effects are not driving much variation in urban performance. The vast majority of urban areas already have good broadband, so you can’t find any apparent difference that is due to urban having less good coverage. Results are less clear in relation to business rates, house prices, qualifications of local workforce, public employment and demographics, but many of these are hard to test, especially given the data available
We ran 3 qualitative workshops, to hear rural businesses’ views. This confirmed our quantitative findings in relation to communications infrastructure. We also heard about various concerns relating to the planning system, and difficulties due to employment law.
May want to cover next steps re. publication – could say aiming for publishing on Defra’s R&D portal in March: http://randd.defra.gov.uk/