English cities and regions: the importance of migration in patterns of growth or decline

1. English cities and regions: the importance of migration in patterns of growth or decline Acknowledgements CURDS colleagues: Tony Champion Simon Raybould Colin Wymer Funding: *preliminary results, not to be quoted* Joseph Rowntree Foundation (Census Programme) Office of the Deputy Prime Minister (New Horizons)

2. Structure of the talk • Why is migration important? • What are the main trends in recent years? • How far can Census data give the details? • Which direction did people move in 2000-1? • Where in London did people move to/from?

3. Why is migration important? At the regional scale Migration is classically the mechanism for labour supply to adjust its location to the distribution of labour demand (the longer-term & longer-distance version of commuting) BUT the net flow tends to take more highly skilled people and so its effects differ for importing/exporting places: exporting region has joblessness alleviated BUT its skill base for regeneration is eroded importing region has skills shortages reduced BUT its cumulative growth can lead to ‘over-heating’ Within a city region Migration is primarily seen as people adjusting housing etc to job locations and domestic constraints and preferences: some areas become favoured … others abandonment

4. …and as a result… Net migration is key indicator of the relative attractivity of any city (region) as a place to live (and/or work) example: Ireland’s migration turn-around as a Celtic Tiger Migration emphasised – indirectly – by Richard Florida’s thesis that certain types of people drive city growth Policy responding by targeting (selective) net in-migration example: Scotland’s aim to attract ‘new blood’ in future Migration patterns can identify city region boundaries, in the form of housing markets, using regionalisation analyses

5. Housing Market Areas (an initial experiment consult before copying)

6. Factors linked to in- and/or out-migration (Champion et al) • demographic (eg. age structure) • cultural/social (eg. ethnicity aspects) • labour market (eg. employment rate) • housing (eg. house price : income ratio) • environmental (eg. air pollution) • policies (eg. on green belts). [static or change measures, as appropriate]

8. Trends in recent years Micro level Possible decline in mobility due to 2-earner households Student mobility levels continuing to be dramatically high Impact of partnership fluidity and more ‘transitory’ behaviour Macro level Cyclical impacts of ‘regional’ house price differentials Increased international flows (including asylum seekers etc) but possibility that lower proportion aim to settle on arrival

9. Recent net international flows

10. DWP analysis of “A8” inflows

11. Spatial analysis of A8 migrants (consult before copying)

12. E.Europe-born residents in 2001

13. What the 2001 Census can tell us • Nothing about migration flows to other countries: this means it is not possible to consider net migration flows across the UK boundary, when some models of housing markets emphasise ‘displacement’ effects • Something about the recent moves of people in different occupation (etc) groups BUT not whether their status changed with the move (eg. from being a student) … nor can the movers have a migration rate calculated because there are no equivalent ‘population’ counts! • Nothing consistent about change in migration patterns since 1990-1 because Census definitions changed ESPECIALLY the Census night location of students

14. Analysis of cities & city regions Cities are taken to be continuously built-up areas which are also economically separable * the starting points are the 2001 Urban Area definitions * each city is the bulk of at least one Travel-to-Work Area * all are also analysed in the State of the Cities research * the largest 24 such cities in England are examined here City regions are as defined by CURDS in the 1990s * each city had most of the 16 features of a regional city * each region is a group of Localities round 1 regional city * commuting/migration flows link the Localities in a region

15. City regions and retail centres

16. Age factor in migration is critical

17. “in/out” ratio: city or city region?

18. Key group: “HM&P MGRPs” [sic: Higher Professional/Managerial]

20. Flows to/from cities beyond their own City Regions • Flows to/from London for the other 26 cities: rates decline with distance from London, but faster for inflows than outflows • In/Out ratios for HMPs higher than 1 only forLondon (plus Brighton and Derby) • London and Reading have ‘perfect’ positive relationship between in/out ratio & skill level • There are ‘perfect’ negatives for Sheffield, Portsmouth, Nottingham, Glasgow, Coventry, Bradford

21. Exploring contrasts by city type • Too few cities for regression, so group them • Analysis of international migration reveals high inflow from abroad for a ‘Gateway’ type of city: London / Reading / Brighton • Examining other cities’ in:out ratios (for all migrants and for HM&Ps) to/from beyond their CRs produces three types which (very crudely) summarised as • Net Gain • Balanced • Net Loss

22. Flows beyond CR boundaries

23. Labour market factors loom large

24. Flows within cities’ own CRs

25. Labour market factors less strong

26. Flows within the London CR

27. The London CR analysis zones

28. Largest flows are within the zones

29. Linkage measure: “Tij squared” Adds together measures on flows Tij + Tji … For flow Tij [ie. the no. migrating zone i to j ] Tij as a proportion of all who migrated from i multiplied by Tij as a proportion of all who migrated to j For flow Tji [ie. the no. migrating zone j to i ] Tji as a proportion of all who migrated from j multiplied by Tji as a proportion of all who migrated to i

30. Tij2 on HM&Ps flows in London

31. Summary from Tij2analyses • Most strong linkages are between adjacent zones, due to distance deterrence • There is evidence that transport axes shape radial linkage patterns • Few radial linkages between conurbation and more rural zones are strong • Linkages are weak across the Thames in the centre/east • London seems to be distinct* in having less difference between the migration patterns of HM&P residents and others in its population * comparative analyses on Birmingham and Bristol CRs

32. In/out ratios: zones and rest of CR

33. …and for HM&Ps only

34. In/out ratios: flows beyond CR

35. …and for HM&Ps only

36. Do intra-CR flows raise diversity?Correlations (r) of zones’ in/out ratios by social group vs. % residents in group

37. Models of net flows: big residuals

38. Main findings from new analyses • Patterns across the zones in the London CR: Broadly centrifugal intra-CR flows, but inflows to centre; young and HM&P have net inward flow within CR too • Relationship between in/out ratio and social profile: Low skill group’s intra-CR flows reinforce social divisions, but HM&P flows neutral; flows from beyond CR strongly reinforce concentrations of HM&Ps in inner areas • Effects of flows for other cities: Inflow of young a key potential asset for city like Newcastle but student retention follows sustained economic growth; any flow increase ‘negative’ for city like Hull (but diversity?) • Speculation on London implications of flows/trends: London growth relies on continued flow of young/gifted who accept poorer quality housing/life to gain ‘escalator region’ career boost: now more likely short-term and from abroad?