Defining functional areas for policy development and implementation
Submitting InstitutionNewcastle University
Unit of AssessmentGeography, Environmental Studies and Archaeology
Summary Impact TypeEconomic
Research Subject Area(s)
Mathematical Sciences: Statistics
Economics: Applied Economics
Summary of the impact
Coombes' research to advance spatial-analysis methodology has re-defined
Travel-to-Work Areas (TTWAs) — the only official UK boundaries defined by
academics — and produced three distinct strands of impact.
Use of Boundaries: TTWA boundaries are widely used by
the UK government and others because they accurately map economic
geography; they are used to select areas for major funding support.
Use of Concept: TTWAs are cited as `bench-mark'
functional economic area definitions in the guidance for implementing UK
policies such as that on Local Economic Partnerships (LEPs).
Use of Methodology: Countries in three continents have
adapted the TTWA method to define functional areas for their official
The need for appropriately-defined areas for policy development and
Many policy decisions rely on comparing statistics on areas, and so using
appropriately-defined areas makes for better policy decisions. An area can
miss out on large sums of public funds if the boundary around that area is
not drawn appropriately. Prior to the definition of TTWAs, all the local
official statistics in the UK were reported for administrative areas, but
administrative areas have idiosyncratic boundaries and this distorts the
comparison of areas. (This sensitivity of statistical analyses to the
areas used is the Modifiable Areal Unit Problem.)
CURDS research activity
Research in the Centre for Urban and Regional Development Studies (CURDS)
on functional area definitions (e.g. TTWAs) had to overcome three
- re-interpreting the local labour market area concept — previously
assumed to be urban-centred — because modern commuting behaviour became
very varied as jobs moved to different areas; the more flexible concept
that was devised is of labour markets as commuting flow clusters, whose
spatial form is not pre-determined;
- creating computer algorithms to implement this concept in analysing
data on patterns of flows, so hugely increased computing power could
utilise the greatly increased detail in available data (e.g. commuting
matrices potentially having 1.7 billion cells);
- being sensitive to the constraints on innovation in official
statistics by adopting a co-production of knowledge approach with
Research mainly supporting impacts through Use of Boundaries
To make the TTWA definition methodology as self-optimising as possible
required significant algorithmic and software innovation, while the
multi-million cell commuting flow matrices initially had to be analysed
with the computing power of the 1990s (1). Census 2001 commuting
datasets then provided unprecedented levels of spatial detail, supporting
major refinements of the TTWA methodology to maximise its potential for
self-optimisation (2). This new refined version of the method was
used to define 2001-based TTWAs as part of co-production of knowledge with
ONS (Office for National Statistics) (3).
Research mainly supporting the Use of Concept
Increasing the flexibility of the TTWA methodology extended its potential
use beyond labour market area definitions, at a time when various new UK
policies were aimed at functional economic areas. One radical use of the
TTWA method applied it to innovative synthetic data (i.e. a dataset
collating diverse patterns of local interactions), thus portraying each
British locality as a `space of flows' (4). The increased
flexibility of the latest version of the methodology further enhanced its
transferability, as exemplified by its application to migration data
analyses to define housing market areas (5).
Research mainly supporting the Use of Methodology
Another form of transferability involves being usable in the very
different geographical conditions, and with the different data available,
in other countries with a recognised need for functional areas. TTWA
definitions set new standards in regionalisation methods, prompting a
review for the US Census Bureau to conclude that `British geographers have
developed a more sophisticated computer algorithm for dividing the country
into labour market areas' (Frey & Speare, cited in (2)).
Applying the method in Italy was a co-production of knowledge with its
national statistical agency, and the resulting "customised" method is
still used there. CURDS research for Eurostat has proved the methodology's
transferability to highly contrasting countries — for example, revealing
the importance of cross-border flows for a country such as Belgium — while
different challenges to the transferability of the method were explored
with respect to countries with little available data (6).
Research all led by Mike Coombes (Professor of Geographic Information by
1998, in CURDS from 1977) working with Colin Wymer (Visiting Fellow,
full-time academic-related staff until 2002).
References to the research
1) Office of National Statistics & Coombes M (1998) 1991-based
Travel-to-Work Areas, ONS, London [co-production of knowledge; CURDS
developed and applied all the analyses]. Available from HEI on request.
2) Coombes M (2010) `Defining labour market areas by analysing commuting
data: innovative methods in the 2007 review of Travel-to-Work Areas'
pp.227-241 in J Stillwell, O Duke-Williams and A Dennett (eds.) Technologies
for Migration and Population Analysis: Spatial Interaction Data
Applications IGI Global, Hershey, PA (USA) DOI:
3) Coombes M and Bond S (2008) Travel-to-Work Areas: The 2007 Review
ONS, London [co-production of knowledge; CURDS developed and applied all
the analyses]. Available from HEI on request.
4) Coombes M (2000) `Defining locality boundaries with synthetic data' Environment
& Planning A 32:1499-1518 DOI 10.1068/929165.
5) Jones C, Coombes M, Dunse N, Watkins D and Wymer C (2012) `Tiered
housing markets and their relationship to labour market areas' Urban
Studies 49:2633-2650 [consortium project; CURDS developed and
applied all the analyses]. REF2 output: 180473.
6) Coombes M (2004) `Multiple dimensions of settlement systems: coping
with complexity' pp.307-324 in A Champion and G Hugo (eds.) New Forms
of Urbanization: Beyond the Urban-Rural Dichotomy Ashgate,
Aldershot ISBN-10: 0754635880. Available from HEI on request.
Key research grants and consultancies:
||Study on Comparable Labour Market Areas
||Eurostat (sub-contract via DevStat)
||Dec 2011-Oct 2012
||The Geography of Housing Market Areas in England
||National Housing & Planning Advisory Unit
||Review of Travel to Work Areas
||Office for National Statistics
||Nov 2006-Dec 2007
||The Economic Viability & Self-Containment of
Geographical Economics: A Framework for Analysis
||Office of the Deputy Prime Minister
||Feb to Oct 2005
||Provision of Intelligence Related to the Definition
of Travel to Work Areas
||Office for National Statistics
||Localities and City Regions Multi Source Definition
for Census Research
Details of the impact
1. Use of Boundaries TTWA boundaries are widely used
by the UK government and others because they accurately map economic
geography: they are used to select areas for major funding support
Numerous series of Official Statistics are reported by TTWAs, some on a
monthly basis, and they are very widely used by labour market analysts and
planners: for example the ONS's Nomis database supports several thousand
TTWA-based downloads each year (IMP1). This usage level reflects
the widespread recognition of TTWAs as the standard UK functional economic
geography, officially endorsed by ONS following research studies when they
"worked with Newcastle University to refine and then apply a complex
allocation process which guarantees that all the TTWAs meet precise
statistical criteria" (IMP2).
Government policy use of TTWAs includes guidance in the Treasury Green
Book on how to assess the case for government intervention: TTWAs are
cited in explaining how analyses should use areas that approximate local
labour markets, and worked examples are given for one application, the
assessment of proposed transport investments in Scotland (IMP3).
Major policies have for a decade now focused on cities, evidenced by the
previous government's State of the English Cities research which relied on
TTWAs. Recent significant developments include the £1.4bn Regional Growth
Fund whose guidance on assessing bids cites TTWAs, whilst the announcement
of City Deals based its choice of cities on TTWAs, even though the policy
is explicitly aimed at local authorities whose boundaries are very
different to TTWAs (IMP4).
These official uses of TTWAs are matched by third sector researchers
utilising them to critique government policy. Examples include analyses by
NESTA of creative clusters, and regional pay policy research finding TTWAs
provide a "more precise way of looking at geographical differences" (IMP5).
Parallel private sector TTWA usage includes their adoption as default
definitions of labour market areas in modelling work, so that if TTWAs did
not exist such work would be "done less well and less efficiently, to the
detriment of both public policy-making and the public purse" (IMP6).
2. Use of Concept TTWAs are cited as the `bench-mark'
functional economic area definitions, as in guidance for implementing UK
policies such as that on Local Economic Partnerships (LEPs)
The coalition government's new policy on `Local Growth' implemented part
of its predecessor's strategy by creating Local Enterprise Partnerships
(LEPs) which straddle administrative boundaries to cover functional
economic areas. The policy documents repeatedly reference TTWAs as the
established template for assessing the functional reality of boundaries
for proposed LEPs (IMP7).
An explicit link between economic development and housing policies led to
a parallel need for housing market area (HMA) definitions which were also
functional areas that potentially cut across administrative boundaries.
Innovative research for the National Housing and Planning Advice Unit
extended the TTWA definitions to identify consistently-defined HMAs, and
despite the coalition government changing housing policy it stated that
"[t]hese areas will help local planning" (IMP8). An academic paper
on the implications of the research for planning professionals won the
2013 RTPI Education Award: Excellence in Spatial Planning.
3. Use of Methodology Countries in three continents
have adapted the TTWA method to define functional areas for their
International organisations like OECD acknowledge that better spatial
policy decisions are made when appropriate areas are used, and for
economic development this means using functional economic areas. A review
for Eurostat has shown that many countries define labour market areas, but
many rely on methods with severe limitations (e.g. when faced with vast
sparse flow matrices — such as those on the c.30,000 French Communes —
they would yield sub-optimal results because of limiting the possible
solutions). Eurostat is convening a Task Force to consider taking forward
the approach to multi-national LMA definitions developed in research in
which "the contribution of Prof Coombes and his colleagues in Newcastle
was substantively innovative and critical for the achievement of the
quality of the study" (IMP9).
The transferability of the method to other countries means it can reveal
the latent spatial structure within commuting datasets on countries with
radically contrasting economic geographies and levels of urbanisation
(e.g. Italy and New Zealand). The results have proved suitable for use
with official statistics, and the increasingly available small area
commuting data in 2010/2011 Censuses is seeing official TTWA-derived
analyses in several countries including South Korea and Canada (IMP10).
Sources to corroborate the impact
(IMP1) Statement from Nomis [this is direct evidence of the usage of
statistics reported by TTWA]. Available on request.
(IMP2) Statement from Office for National Statistics [this is evidence of
the several co-production of knowledge projects by ONS and researchers at
Newcastle, and of their value to ONS]. Available on request.
(IMP3) Transport Scotland (2008) Developing a Case by Case Approach
The Scottish Government, Edinburgh www.transportscotland.gov.uk/stag/td/Part2/Economy/220.127.116.11
[there is a worked example of the Treasury guidance using TTWAs in Tables
9.3 & 9.5].
(IMP4) HM Government (2011) Unlocking Growth in Cities Cabinet
Office, London www.gov.uk/government/uploads/system/uploads/attachment_data/file/7523/CO_Unlocking_20GrowthCities_acc.pdf
[the quote is on page 3, showing
public funding guided by TTWA]
(IMP5) Incomes Data Services (2011) Location-Based Pay
Differentiation Unison, London www.docstoc.com/docs/158456102/IDS-research-paper-for-UNISON-FINAL-16-09-11-_2_
[the quote is from page 26 and is evidence of the
recognition of the robustness of TTWAs].
(IMP6) Letter from Private Sector Consultancy [this letter evidences the
use of TTWAs in the private sector, and its value to their clients, who
are often policymakers in the public sector]. Available on request.
(IMP7) HM Government (2010) Local Enterprise Partnerships Letter
from Secretaries of State BIS & CLG to local authority and business
[specific reference to travel to work areas is on page 2; White Papers and
documentation laying the foundations for Local Enterprise Partnerships had
cited TTWAs as a bench-mark].
(IMP8) Department of Communities and Local Government (2010) Geography
of Housing Market Areas: Final Report and Summary www.gov.uk/government/publications/housing-market-areas
[the whole document describes the development and application of a version
of the TTWA method which applies its principles to the definition of
Housing Market Areas].
(IMP9) Statement from Eurostat [this is evidence of the key role of work
in Newcastle as the basis for the potential development of harmonised LMA
definitions covering the whole of the EU]. Available on request.
(IMP10) Munro A, Alasia A, Bollman R (2011) Self-Contained Labour
Areas: A Proposed Delineation and Classification by Degree of Rurality
Statistics Canada, Ottowa publications.gc.ca/collections/collection_2011/statcan/21-006-X/21-006-x2008008-eng.pdf
[the whole document describes the application of an adapted version of the