edge document template - northumberland...1.3 since the september 2014 forecasts were produced, new...

Northumberland

County-Level Demographic Analysis &

Forecasts

July 2015

Leeds Innovation Centre | 103 Clarendon Road | Leeds | LS2 9DF

0113 384 6087 | www.edgeanalytics.co.uk

For the attention of:

Steve Robson

Senior Housing Policy Officer

Northumberland County Council

i

July 2015

Acknowledgements

Demographic statistics used in this report have been derived from data from the Office for

National Statistics licensed under the Open Government Licence v.1.0.

The authors of this report do not accept liability for any costs or consequential loss involved following the use of the data and analysis referred to here; this is entirely the responsibility of the users of the information presented in this report.

ii

July 2015

Table of Contents

Acknowledgements ................................................................................................................. i

Table of Contents ....................................................................................................................ii

1 Introduction .................................................................................................................... 1

2 Area Profile: Northumberland ......................................................................................... 5

3 Scenario Development ................................................................................................... 16

4 Scenario Outcomes......................................................................................................... 25

5 Summary ....................................................................................................................... 34

POPGROUP Methodology ................................................................................. 37 Appendix A

Data Inputs & Assumptions .............................................................................. 40 Appendix B

1

July 2015

1 Introduction

Context

In July 2014, Edge Analytics was commissioned by Northumberland County Council (NCC) to 1.1

provide a range of demographic forecasts as part of the on-going work for the emerging

Northumberland Local Plan.

The demographic scenarios were developed at both county and small-area (i.e. sub-county) level 1.2

and included the most recent 2012-based sub national population projections (SNPP) from the

Office for National Statistics (ONS) as the official benchmark scenario. A number of ‘dwelling-led’

scenarios were also produced, in which population growth was determined by growth in the

number of dwellings. Additionally, ‘jobs-led’ scenarios were developed at county-level to assess

the level of population growth associated with the employment growth trajectory implied by the

St Chad’s employment forecasts1. Forecasts were provided for the 2011–2031 Plan Period, with

household growth assessed using assumptions from both the 2008-based and 2011-based

interim household projection models from the Department for Communities and Local

Government (DCLG)

Since the September 2014 forecasts were produced, new demographic statistics have become 1.3

available. In June 2014, the ONS released the 2013 mid-year population estimate, providing an

additional year of historical population data for use in the development of trend-based

demographic scenarios. In February/March 2015, the DCLG released the 2012-based household

projection model, superseding the 2011-based interim model.

1 These were referenced in the report entitled ‘Long Term Sectoral, Employment and Land Use Projections’ Policy Research Group, St Chad’s College, Durham.

2

July 2015

Requirements

NCC has requested that Edge Analytics produce an updated range of demographics forecasts at 1.4

both county level and small-area level, using the latest demographic statistics. These forecasts

use the latest household-growth assumptions from the 2012-based DCLG household projection

model.

This project is to be provided in two phases: Phase 1 will provide a range of scenarios at county 1.5

level whilst Phase 2 will include demographic forecast(s) at sub-county (i.e. small-area) level.

NCC has also requested that a range ‘sensitivity’ scenarios are developed, to evaluate the impact 1.6

of alternative commuting and unemployment assumptions on the population and dwelling

growth outcomes of the jobs-led scenarios.

Additionally, NCC has requested that a migration sensitivity scenario be developed to consider 1.7

the impact of increased in-migration from North Tyneside to Northumberland, for consistency

with a scenario considered by North Tyneside Council. Other neighbouring authorities have not

made explicit assumptions about Northumberland in relation to migration.

Approach

Official Guidelines

The development and presentation of demographic evidence to support local housing plans is 1.8

subject to an increasing degree of public scrutiny. The National Planning Policy Framework

(NPPF)2 and Planning Practice Guidance (PPG)3 provide guidance on the appropriate approach to

the objective assessment of housing need. The PPG states that the DCLG household projections

should provide the “starting point estimate of overall housing need” (PPG paragraph 2a-015).

Local circumstances, alternative assumptions and the most recent demographic evidence,

including ONS population estimates, should also be considered (PPG paragraph 2a-017). Evidence

that links demographic change to forecasts of economic growth should also be assessed (PPG

paragraph 2a-018).

2 http://planningguidance.planningportal.gov.uk/blog/policy/

3 http://planningguidance.planningportal.gov.uk/blog/guidance/

http://planningguidance.planningportal.gov.uk/blog/policy/

http://planningguidance.planningportal.gov.uk/blog/guidance/

3

July 2015

The use of demographic models, which enable a range of growth scenarios to be evaluated, is 1.9

now a key component of the objective assessment process. The POPGROUP suite of demographic

models, which is widely used by local authorities and planners across the UK, provides a robust

and appropriate forecasting methodology (for further information on POPGROUP, refer to

Appendix B ).

The choice of assumptions used within POPGROUP has an important bearing on scenario 1.10

outcomes. This is particularly the case when trend projections are considered alongside

population and household forecasts. The scrutiny of demographic assumptions is now a critical

component of the public inspection process, providing much of the debate around the

appropriateness of a particular objective assessment of housing need.

Edge Analytics’ Approach

Edge Analytics has used POPGROUP v.4 technology to develop a range of demographic scenarios 1.11

for Northumberland. As the starting point of this assessment, the most recent official population

and household projections are considered. The 2012-based SNPP for Northumberland is

presented, with an analysis of the ‘components of change’ underlying this new projection. These

statistics are compared to previous estimates and to the historical data on births, deaths and

migration. The most recent 2012-based DCLG household projection model is also considered,

with commentary provided on the differences between this and the earlier household projection

models.

In line with the PPG, Edge Analytics has developed a range of demographic scenarios for 1.12

Northumberland, with the 2012-based SNPP presented as the official ‘benchmark’ scenario. A

number of alternative scenarios have been developed using the latest demographic and

economic assumptions, for comparison with the 2012-based SNPP. The alternative scenarios

include ‘trend’ scenarios, based on varying migration assumptions and ‘jobs-led’ scenarios, in

which population change is determined by growth in the number of jobs. Dwelling-led scenarios

have also been developed, in which population growth is determined by the growth in the

number of dwellings.

Sensitivity analysis has been conducted to evaluate the impact of varying economic assumptions 1.13

on the growth-outcomes of the jobs-led scenarios. A migration sensitivity scenario has also been

4

July 2015

developed, to assess the demographic implications of increased in-migration from North

Tyneside.

In line with the PPG, the household growth implications of each scenario are assessed using 1.14

assumptions from the 2012-based DCLG household projection model. As outlined in the PPG, it is

appropriate to consider “alternative assumptions in relation to the underlying demographic

projections and household formation rates” of the local area (PPG Paragraph 2a-017). For

comparison, household growth has therefore also been assessed using assumptions from the

earlier 2008-based DCLG household projection model. This household projection model was

calibrated under very different market conditions to the 2012-based model, and consequently

shows higher rates of household formation.

All scenarios have been run to a 2031 horizon, with historical data included for the 2001–2013 1.15

period. Scenario results are presented for the 2011–2031 plan period.

Report Structure

The report is structured in the following way: 1.16

In Section 2, a profile of Northumberland is presented. This includes a historical

perspective on population change since the 2001 Census, analysis of the ‘components

of change’ from the 2012-based SNPP, and commentary on the 2012-based DCLG

household projection model.

In Section 3, a definition of each scenario is presented, with the outcomes of these

scenarios detailed in Section 4.

Section 5 summarises the analysis and identifies a number of key issues for

Northumberland County Council to consider.

Appendix A presents an overview of the POPGROUP methodology.

Appendix B provides detail on the data inputs and assumptions used in the

development of the POPGROUP scenarios.

5

July 2015

2 Area Profile: Northumberland

Introduction

The development of local plans is made considerably more challenging by the dynamic nature of 2.1

key data inputs. Economic and demographic factors, coupled with the continuous release of new

statistics, often undermine the robustness of underpinning evidence. This has been a particular

issue since 2012, with the release of 2011 Census statistics, revisions to historical population

estimates and updated population and household projections.

This section provides an overview of population change in Northumberland since 2001 and the 2.2

recent revisions to the mid-year population estimates. Also presented is the most recent

population projection from ONS, the 2012-based SNPP and its constituent ‘components of

change’. Commentary is also provided on the most recent household projections, the 2012-based

household projection model from DCLG.

Population Change 2001–2011

Mid-Year Population Estimates

Between successive Censuses, population estimation is necessary. These mid-year population 2.3

estimates (MYEs) are derived by applying the ‘components of change’ (i.e. counts of births and

deaths and estimates of internal and international migration) to the previous year’s MYE.

Following the 2011 Census, the 2002–2010 MYEs were ‘rebased’ to align them with the 2011

MYE and to ensure the correct transition of the age profile of the population over the 2001–2011

decade.

At the 2011 Census, the resident population of Northumberland was 316,028, a 2.8% increase 2.4

over the 2001–2011 decade. The 2011 Census population total proved to be higher than that

6

July 2015

suggested by the trajectory of growth from the previous MYEs. For this reason, the revised final

MYEs are higher than the previous MYEs (Figure 1).

Figure 1: Northumberland Mid-Year Population Estimates (Source: ONS)

Components of Change

The rebasing of the MYEs involved the recalibration of the components of change for 2001/02–2.5

2010/11. Between Censuses, births and deaths are accurately recorded in vital statistics registers

and provide a robust measure of ‘natural change’ (the difference between births and deaths) in a

geographical area. Given that births and deaths are robustly recorded, and assuming that the

2001 Census provided a robust population count, the 'error' in the MYEs is due to the difficulties

associated with the estimation of migration.

Internal migration (i.e. migration flows to and from other areas in the UK) is adequately 2.6

measured using data from the Patient Register (PR), the National Health Service Central Register

(NHSCR) and Higher Education Statistics Agency (HESA), although data robustness may be lower

where there is under-registration in certain age-groups (young males in particular). It is therefore

most likely that the ‘error’ in the previous MYEs is associated with the mis-estimation of

international migration, i.e. the balance between immigration and emigration flows to and from

Northumberland.

However, ONS has not explicitly assigned the MYE adjustment to international migration. Instead 2.7

it has identified an additional ‘unattributable population change’ (UPC) component, suggesting it

has not been able to accurately identify the source of the 2001–2011 over-count (Figure 2). The

7

July 2015

effect of the UPC adjustment depends upon the scale of population recalibration that has been

required following the 2011 Census results. For Northumberland, the population estimates have

been subject to a consistent annual increase due to the under-count over the 2001–2011 decade.

Figure 2: Northumberland Mid-Year Population Estimates (Source: ONS)

For demographic analysis, the classification of UPC is unhelpful, but given the robustness of 2.8

births, deaths and internal migration statistics compared to international migration estimates, it

is assumed that it is most likely to be associated with the latter. With the assumption that the

UPC element is assigned to international migration (for estimates up to 2011), and with the

inclusion of statistics from the 2012 and 2013 MYEs from ONS, a twelve-year profile of the

‘components of change’ is presented for Northumberland (Figure 3).

Figure 3: Northumberland components of population change 2001/02 to 2012/13, including the UPC component in the 2001/02 to 2010/11 international migration component. (Source: ONS).

8

July 2015

For Northumberland, the positive population change over the 2001/02–2012/13 period was 2.9

predominantly driven by net internal migration. With the exception of 2011/12 and 2012/13, net

international migration has been largely positive throughout the historical period. Natural change

remained negative throughout the historical period, with the number of deaths exceeding the

number of births.

Internal Migration Flows

As illustrated in Figure 3 net internal migration has been a dominant component of population 2.10

change over the 2001/02–2012/13 historical period. Internal in-migration has been consistently

higher than internal out-migration over the historical period (Figure 4E), resulting in a net

increase in the population.

The neighbouring authorities of Newcastle, North Tyneside and Gateshead have historically had 2.11

the highest inflows and outflows to and from Northumberland (Figure 4A and B). In terms of net

migration, the greatest net inflows are from these three authorities (Figure 4C), with the greatest

net outflows to Leeds, Manchester and Carlisle (Figure 4D).

Net internal migration was positive in the majority of age groups in Northumberland (Figure 4F) 2.12

over the 2001/02–2012/13 historical period. Only the 15–19 age group experienced a net

outflow of migrants from Northumberland, this is usually students leaving to go to university.

The internal in- and out-flows between Northumberland and its neighbouring authorities over 2.13

the 2001/02–2012/13 period are summarised in Figure 5.

9

July 2015

Figure 4: Historical internal migration profile for Northumberland 2001/02–2012/13 (Source: ONS)

Average 2001/02-2012/13 Average 2001/02-2012/13

Newcastle upon Tyne 1,348 Newcastle upon Tyne 1,613

North Tyneside 1,058 North Tyneside 1,320

Gateshead 509 Gateshead 607

County Durham UA 469 County Durham UA 454

Leeds 227 Sunderland 237

Sunderland 223 Leeds 194

South Tyneside 129 South Tyneside 163

Carlisle 125 Carlisle 105

Sheffield 93 Sheffield 79

Manchester 89 York 77

(B) Top Ten Inflows(A) Top Ten Outflows

Northumberland: Internal Migration Profile

266

263

98

34

26

14

10

10

9

9

0 100 200 300 400

Newcastle upon Tyne

North Tyneside

Gateshead

South Tyneside

Bradford

Sunderland

Bromley

St. Albans

Ealing

Stockport

Average Net Migration 2001/02 - 2012/13

(D) Top Ten Net Inflows

-9

-9

-11

-12

-14

-14

-16

-21

-25

-33

-40 -30 -20 -10 0

North Lincolnshire

Oxford

York

Nottingham

Sheffield

Middlesbrough

County Durham UA

Carlisle

Manchester

Leeds

Average Net Migration 2001/02–2012/13

(C) Top Ten Net Outflows

0

2,000

4,000

6,000

8,000

10,000

12,000

200

1/0

2

200

2/0

3

200

3/0

4

200

4/0

5

200

5/0

6

200

6/0

7

200

7/0

8

200

8/0

9

200

9/1

0

201

0/1

1

201

1/1

2

201

2/1

3

Inte

rnal

Mig

ratio

n

(E) Inflows, Outflows & Net Flows

Net In Out

-1000

-800

-600

-400

-200

0

200

400

0-4

5-9

10-1

4

15

-19

20-2

4

25

-29

30-3

4

35

-39

40-4

4

45

-49

50-5

4

55

-59

60-6

4

65

-69

70-7

4

75

+

Ne

t M

igra

tio

n A

v 2

00

1/0

2-2

01

2/1

3

(F) Age Group Net Flows

10

July 2015

Figure 5: Historical internal migration flows between Northumberland and the neighbouring local authorities 2001/02–2012/13 (source: ONS). Green indicates the flows from the relevant neighbouring authority to Northumberland and red the flows from Northumberland.

11

July 2015

Official Projections

Official Population Projections

In the development and analysis of population forecasts, it is important to benchmark any 2.14

growth alternatives against the latest ‘official’ population projection. The most recent official

subnational population projection is the ONS 2012-based SNPP, released in May 2014.

SNPPs are released every couple of years by the ONS. These projections are trend-based and 2.15

provide an indication of population growth over a 25-year period. In 2011, the ONS published the

2011-based interim SNPP, in which population growth was projected over a shorter 10-year

period (2011–2021). In May 2014, the 2012-based SNPP was released, providing a new

‘benchmark’ for the analysis of population growth.

Under the 2012-based SNPP, the population of Northumberland is expected to increase by 8,630 2.16

over the 2012–2037 forecast period, a 2.7% increase. This is lower than under the previous

population projection, the 2010-based SNPP, in which population growth was projected to be

4.7% over the 2010–2035 projection period (Figure 6).

Figure 6: Official ONS population projections (source: ONS)

12

July 2015

The 2012-based SNPP components of change are presented in Figure 7, with the historical 2.17

components of change for 2001/02 to 2011/12 included for comparison. The annual average

natural change, net migration (internal and international) and population change for the

2012-based SNPP are compared to the historical 5-year and 10-year averages in Table 1.

Figure 7: Historical and 2012-based SNPP components of change for Northumberland (source: ONS)

Table 1: 2012-based SNPP components of change (source: ONS)

Historically, over both the 5- and 10-year periods, net internal migration has been the dominant 2.18

component of population change (+489 and +784 per year respectively). Under the 2012-based

SNPP, net internal migration is projected to be higher, averaging +1,224 per year over the

2012/13–2036/37 period.

Natural Change

Net Internal Migration

Net International Migration

Unattributable Population Change*

Annual Population Change

Annual Population Change (%)

* UPC is only applicable to the years 2001/02 - 2010/11

345

0.11%

Component of Change

611

0.20%

-361

784

-41

350

752

0.25%

Historical Projected

-221

489

-24

351

-773

10-year average

(2002/03–2011/12)

2012-based SNPP

average

(2012/13–2036/37)

1,224

-106

-

5-year average

(2007/08–2011/12)

13

July 2015

Natural change has been negative over the historical 5- and 10-year periods, with the number of 2.19

deaths exceeding the number of births. Under the 2012-based SNPP, the number of deaths is

expected to continue to exceed the number of births, but at a higher rate than that seen

historically. Net international migration has been negative historically; this trend is continued in

the 2012-based SNPP, but at a higher rate than over both the 5- and 10-year periods.

Official Household Projections

In the assessment of housing need, the PPG states that the latest DCLG household projections 2.20

should provide the starting point estimate (PPG paragraph 2a-015). The latest 2012-based

household projection model, which is underpinned by the 2012-based SNPP, was released by the

DCLG in February/March 2015, superseding the 2011-based interim household projection model.

The methodological basis of the new 2012-based model is consistent with that employed in the 2.21

previous 2011-based interim and 2008-based household projections. A ‘two-stage’ methodology

has been used by DCLG. ‘Stage One’ produces the national and local projections for the total

number of households by age-group and relationship status group over the projection period.

‘Stage Two’ provides the detailed household type breakdown by age. Currently, only Stage One

output is available for the 2012-based household projection model (refer to Appendix B for

further detail).

Whilst methodologically similar to previous releases, the 2012-based household projections 2.22

provide an important update on the 2011-based interim household projections with the inclusion

of the following information:

2012-based SNPP by sex and age that extend to 2037 (rather than to 2021 as was the

case in the 2011-based interim projections).

Household population by sex, age and relationship-status consistent with the 2011

Census (rather than estimates for 2011, which were derived from 2001 Census data,

projections and national trends, as used in the 2011-interim projections).

Communal population statistics by age and sex consistent with the 2011 Census

(rather than the previous estimate, which were calibrated to the total communal

population from the 2011 Census).

Further information on household representatives from the 2011 Census relating to

aggregate household representative rates by relationship status and age.

14

July 2015

Aggregate household representative rates at local authority level, controlled to the

national rate, based on the total number of households divided by the total adult

household population (rather than the total number of households divided to the total

household population).

Adjustments to the projections of the household representative rates in 2012 based on

the Labour Force Survey (LFS).

(Source: DCLG Methodology4)

Household Model Comparison

In this section, the latest 2012-based household projection model is compared to the earlier 2.23

2008-based model. The 2012-based and 2008-based household projection models are

underpinned by the 2012-based SNPP and 2008-based SNPPs respectively.

Under the 2012-based household projection model, the number of households is expected to 2.24

increase by +14,500 over the 2012–2037 projection period (+580 households per year). Under

the previous 2008-based household projection model, the number of households was expected

to increase by +24,500 (+980 per year) over the 2008–2033 projection period (Figure 8).

Figure 8: Average number of household under the 2008-based and 2012-based DCLG household projection models (Source: DCLG)

Average household size is slightly lower under the 2012-based model than under the 2008-based 2.25

model, although both models suggest similar reductions in household size over their relevant

4 Household Projections 2012-based: Methodological Report. Department for Communities and Local Government (February 2015). https://www.gov.uk/government/statistics/2012-based-household-projections-methodology

https://www.gov.uk/government/statistics/2012-based-household-projections-methodology


15

July 2015

projection periods: the 2012-based model suggests average household size will reduce from 2.23

to 2.06 2012–2037, whereas the 2008-based model suggested a reduction from 2.24 to 2.05

(2008–2033) (Figure 9).

Figure 9: Average household size under the 2008-based and 2012-based DCLG household projection models (Source: DCLG)

16

July 2015

3 Scenario Development

Introduction

There is no single definitive view on the likely level of growth expected in Northumberland; a mix 3.1

of economic, demographic and national/local policy issues ultimately determines the speed and

scale of change. For local planning purposes, it is necessary to evaluate a range of growth

alternatives to establish the most ‘appropriate’ basis for determining future housing provision.

Edge Analytics has used POPGROUP (v.4) technology to develop a range of growth scenarios for 3.2

Northumberland (for detail on the POPGROUP methodology, refer to Appendix B).

In line with the PPG, the most recent demographic and household projection models have been 3.3

considered. Nine ‘core’ scenarios have been produced, including: the most recent official

population projection from ONS, the 2012-based SNPP, and two alternative trend-based

scenarios, based on short-term and long-term migration histories. Jobs-led and dwelling-led

scenarios have also been developed, in which population growth is determined by the annual

change in the number of jobs and dwellings respectively.

Jobs-led ‘sensitivity’ scenarios have been developed, in which the demographic implications of 3.4

alternative commuting and unemployment assumptions are assessed. Additionally, a sensitivity

scenario has been developed, which evaluates the impact of increased internal migration flows to

Northumberland.

In the following sections, the scenarios are described and the broad assumptions specified. For 3.5

further detail on the data inputs and assumptions, please refer to Appendix B.

17

July 2015

Core Scenarios

Official Projections

In accordance with the PPG, the alternative scenarios are ‘benchmarked’ against the most recent 3.6

official population projections from the ONS, the 2012-based SNPP. The ‘SNPP-2012’ scenario

replicates this official population projection.

Alternative Trend Scenarios

A five year historical period is a typical time-frame from which migration ‘trend’ assumptions are 3.7

derived (this is consistent with the ONS official methodology). Given the unprecedented

economic change that has occurred since 2008, and the differences between the projected

2012-based SNPP data and the historical data (see paragraph 2.18 on page 12) it is important to

give due consideration to an extended historical time period for assumption derivation.

The following alternative trend scenarios have been developed, based upon the latest 3.8

demographic evidence:

PG-5yr: internal migration rates and international migration flow assumptions are

based on the last five years of historical evidence (2008/09–2012/13), with the UPC

adjustment included within the international migration assumptions.

PG-10yr: internal migration rates and international migration flow assumptions are

based on the last 10 years of historical evidence (2003/04–2012/13), with the UPC

adjustment included within the international migration assumptions.

Labour Force, Jobs-Growth & Jobs-led Scenarios

In the ‘official’ and ‘alternative trend’ scenarios described above, the labour force and jobs-3.9

growth implications of the population growth trajectories have been evaluated using three key

data inputs: economic activity rates, an unemployment rate and a commuting ratio.

In all of the core scenarios, the unemployment rate has been incrementally reduced between 3.10

2014 and 2020 to account for economic recovery post-recession. The economic activity rates

have been adjusted to account for changes to the state pension age (SPA). The 2011 Census

18

July 2015

commuting ratio for Northumberland has been applied, fixed throughout the forecast period (see

Appendix B for further detail on the employment assumptions used).

In a jobs-led scenario, these data inputs are used to determine the level of population growth 3.11

associated with a defined jobs-growth trajectory. Four jobs-led scenarios have been produced,

using jobs-growth figures from the St Chad’s employment forecasts for Northumberland (Figure

10)5:

Jobs-led Baseline: +6,253 jobs 2013/14–2030/31

Jobs-led Lower: +3,559 jobs 2013/14–2030/31

Jobs-led Upper: +11,162 jobs 2013/14–2030/31

Jobs-led Policy: +11,826 jobs 2013/14–2030/31

Figure 10: Jobs-growth trajectory under each of the employment forecast alternatives for

Northumberland (Source: Policy Research Group, St Chad’s College Durham). Note that the jobs-growth targets are applied from the start of the forecast period (i.e. from 2013/14 onwards).

Households, Dwelling-Growth & Dwelling-led Scenarios

In the ‘official’ ‘alternative trend’ and ‘jobs-led’ scenarios, the number of dwellings is derived 3.12

from the population growth trajectory under each of the scenarios using three key assumptions:

headship rates and communal population statistics and a dwelling vacancy rate.

5 Policy Research Group, St Chad’s College Durham

19

July 2015

In a dwelling-led scenario, these data inputs are used to determine the level of population 3.13

growth associated with a defined dwelling-growth trajectory. Two dwelling-led scenarios have

been produced:

Dwelling-led RSS: Population growth is determined by growth in the number of

dwellings, defined using the Regional Spatial Strategy (RSS) target of +15,785 over the

2013/14–2030/31 forecast period.

Dwelling-led Past Delivery: Population growth is determined by growth in the number

of dwellings, defined using the 5-year average completion rate (2009/10–2013/14). In

2013/14 the actual dwelling completion figure is used; from 2014/15 onwards the 5-

year average is applied. This is a total dwelling growth of +11,507 over the 2013/14–

2030/31 forecast period.

Figure 11: Dwelling growth under the RSS requirement and the past 5-years of delivery (Source: Northumberland County Council)

Household & Dwelling-Growth Implications

In all scenarios, household growth has been assessed through the application of household 3.14

headship rates from both the 2012-based and the 2008-based DCLG household projection

models, producing two alternative outcomes for each scenario:

20

July 2015

In the HH-12 outcome, the 2012-based DCLG headship rates are applied;

In the HH-08 outcome, the 2008-based DCLG headship rates are applied, scaled to be

consistent with the 2011 DCLG household total but following the original trend

thereafter.

For further detail on the household-growth assumptions used, please refer to Appendix B. 3.15

Sensitivity Scenarios

Jobs-led Sensitivity Scenarios

In a jobs-led scenario, the economic activity rates, unemployment rate and commuting ratio 3.16

determine the level of population growth associated with the defined jobs-growth trajectory, as

implied by the employment forecasts from the Policy Research Group (St Chad’s College

Durham). To evaluate the sensitivity of the population and dwelling-growth outcomes to changes

in these assumptions, a range of jobs-led sensitivity scenarios have been developed. In these

sensitivities, adjustments have been made to the commuting ratio and unemployment rate

assumptions. These adjustments are reasonable as they based on historical trends and NCC’s

strategy plans.

The commuting ratio determines the balance between the number of jobs in Northumberland 3.17

and the number of resident workers. The 2011 Census Travel to Work statistics suggest a net out-

commute from Northumberland, lower than the net out-commute seen at the 2001 Census (see

Appendix B for further detail). Two jobs-led commuting ratio sensitivities have been developed:

Jobs-led SENS A: The commuting ratio incrementally reduces between 2011 and

2031 at the same rate as between the 2001 and 2011 Censuses: the commuting

ratio incrementally reduces from the 2011 Census value of 1.18 to 1.03 by 2031 (i.e.

a reducing net out-commute).

Jobs-led SENS B: The commuting ratio incrementally reduces between 2011 and

2031 in line with the NOMIS line of best fit, as defined by Northumberland County

Council. The commuting ratio reduces from the 2011 Census value of 1.18 to 1.09

by 2031 (i.e. a reduced net out-commute).

21

July 2015

In the Jobs-led SENS A and B scenarios, the unemployment rate and economic activity rate 3.18

assumptions are consistent with the core scenarios. Two additional jobs-led sensitivity scenarios

have been developed, in which the unemployment rate and the commuting ratio have been

adjusted:

Jobs-led SENS C: The unemployment rate incrementally reduces from the 2014

value of 6.6% to 4.4% by 2031, the lowest historical unemployment rate over the

2004–2014 decade. The commuting ratio reduces from 1.18 in 2011 to 1.03 by 2031

(i.e. consistent with Jobs-led SENS A).

Jobs-led SENS D: The unemployment rate incrementally reduced from the 2014

value of 6.6% to 4.4% by 2031. This is the lowest historical unemployment rate

experienced over the 2004–2014 period. The commuting ratio reduces from 1.18 in

2011 to 1.09 by 2031 (i.e. consistent with Jobs-led SENS B).

In the Jobs-led SENS C and D scenarios, the economic activity rate assumptions are consistent 3.19

with the core scenarios.

Migration Sensitivity Scenario

As highlighted in the Area Profile section, Northumberland has seen a net in-flow of people from 3.20

the surrounding districts of Newcastle, Gateshead and North Tyneside. As part of its 2014 SHMA

preparation, North Tyneside Council considered a scenario that evaluated the impact of a greater

out-flow of migrants to Northumberland6. To reflect the relationship between North Tyneside

and Northumberland, NCC has therefore requested that a similar scenario be produced for

Northumberland.

In the PG-10yr SENS scenario, the internal in-migration flows have been adjusted on the 3.21

assumption that in-migration from North Tyneside to Northumberland will increase in the future.

Internal in-migration is increased at intervals of 50 between 2014/15 and 2020/21. From 2021/22

onwards, internal in-migration is increased at a rate of +200 per year. The PG-10yr scenario has

been used as it assumes the highest forecast growth through migration, compared to the official

SNPP-2012 and the PG-5yr alternative trend based scenario. Additionally, the PG-10yr scenario

has been selected for consistency with the North Tyneside sensitivity scenario.

6 http://www.northtyneside.gov.uk/pls/portal/NTC_PSCM.PSCM_Web.download?p_ID=558934

http://www.northtyneside.gov.uk/pls/portal/NTC_PSCM.PSCM_Web.download?p_ID=558934

22

July 2015

Scenario Summary

Nine core scenarios have been developed under four scenario types; official projections, 3.22

alternative trend, jobs-led and dwelling-led (Table 2). Jobs-led and migration sensitivity scenarios

have been developed in which the impact of alternative economic and internal migration

assumptions are assessed (Table 3). In all scenarios, household growth has been assessed using

assumptions from both the latest 2012-based and the earlier 2008-based household projection

models.

Table 2: Core scenario definition summary

Scenario Type

Scenario Name

Scenario Definition

Official Projection

SNPP-2012 This scenario mirrors the 2012-based SNPP from ONS for

Northumberland. This scenario is the official ‘benchmark’ scenario.

Alternative Trend Based

PG-5yr

Internal and international migration assumptions are based on the

last five years of historical evidence (2008/09 to 2012/13). The UPC

is included in international migration assumptions.

PG-10yr

Internal and international migration assumptions are based on the

last ten years of historical evidence (2003/04 to 2012/13). The UPC

is included in international migration assumptions.

Jobs-led

Jobs-led Baseline

Population growth is determined by the annual change in the

number of jobs, as defined by St Chad’s ‘Baseline’ employment

forecast (headcount totals). This is a total increase of +6,253 jobs

2013/14–2030/31.

The unemployment rate is incrementally reduced over the 2014–

2020 period. The 2011 Census commuting ratio is applied, fixed

throughout the forecast period.

Jobs-led Lower


number of jobs, as defined by St Chad’s ‘Lower’ employment


2013/14–2030/31.




23

July 2015

Scenario Type

Scenario Name

Scenario Definition

Jobs-led Upper


number of jobs, as defined by St Chad’s ‘Upper’ employment


2013/14–2030/31.




Jobs-led Policy


number of jobs, as defined by St Chad’s ‘Policy’ employment


2013/14–2030/31.




Dwelling-led

Dwelling-led RSS


number of dwellings, as defined by the Regional Spatial Strategy

(RSS). This is a total dwelling growth of +15,785 over the 2013/14–

2030/31 period.

Dwelling-led Past Delivery


number of dwellings, based on the historical dwelling completions.

For 2013/14 the known completion rate is used. From 2014/15

onwards a 5-year historical average is used (2009/10–2013/14). This

is a total dwelling growth of +11,507 over the 2013/14–2030/31

period.

Note: Refer to Appendix B for more detail on data inputs and assumptions.

24

July 2015

Table 3: Jobs-led sensitivity scenario definition summary

Sensitivity Type

Scenario Name

Scenario Definition

Jobs-led Sensitivity Scenarios

SENS A

The commuting ratio is incrementally reduced at the same rate as

between the 2001 Census and 2011 Census commuting ratios. In

2011 the Census commuting ratio of 1.18 is applied, incrementally

decreasing to 1.03 by 2031.

Unemployment and economic activity rate assumptions are

consistent with the core scenarios.

SENS B

The commuting ratio is incrementally reduced in line with the

NOMIS line of best fit, as defined by Northumberland County

Council. In 2011 the Census commuting ratio of 1.18 is applied,

incrementally decreasing to 1.09 by 2031.

Unemployment and economic activity rate assumptions are

consistent with the core scenarios.

SENS C

The unemployment rate is incrementally reduced from the 2014

value of 6.6% to the lowest historical unemployment rate of 4.4%

by 2031.

Commuting ratio and economic activity rate assumptions consistent

with the Jobs-led SENS A sensitivity scenarios.

SENS D

The unemployment rate is incrementally reduced from the 2014

value of 6.6% to the lowest historical unemployment rate of 4.4%

by 2031.

Commuting ratio and economic activity rate assumptions consistent

with the Jobs-led SENS B sensitivity scenarios.

Migration PG-10yr SENS

Adjusted internal in-migration flows to reflect increased in-

migration from North Tyneside. Internal migration flows increase by

+50 intervals between 2014/15–2017/18. From 2018/19 onwards

the internal in-migration increases by +200 per year.

Internal out-migration and international migration assumptions are

consistent with the PG-10yr scenario.

Note: Refer to Appendix B for more detail on data inputs and assumptions.

25

July 2015

4 Scenario Outcomes

Introduction

A range of demographic scenarios has been developed for Northumberland using POPGROUP v.4 4.1

technology, with the latest 2012-based SNPP included as the official benchmark scenario.

Scenarios are split into ‘core’ and ‘sensitivity’ scenarios: in this section the scenario outcomes are

presented.

In all scenarios, household growth has been assessed using headship rate assumptions from the 4.2

latest 2012-based DCLG household projection model and the earlier 2008-based model. Scenario

outcomes are therefore presented under an HH-12 and HH-08 alternative.

Core Scenario Outcomes

The core scenario outcomes are presented in the form of two charts and two tables, under an 4.3

HH-12 and an HH-08 alternative. The charts illustrate the trajectory of population growth under

each of the core scenarios (Figure 12 and Figure 13) for the 2001–2031 period, with the historical

mid-year population estimates included for 2001–20137.

The tables summarise the population and household growth outcomes under each of the core 4.4

scenarios, ranked in order of population growth (Table 4 and Table 5). The tables also show the

estimated average annual net migration and dwelling growth associated with the population

change. In all but the dwelling-led scenarios, only the household and dwelling-growth outcomes

differ between the HH-12 and HH-08 outcomes, reflecting the alternative headship rate

assumptions used. In the dwelling-led scenarios, the same dwelling growth targets have been

applied in the HH-12 and HH-08 alternatives from 2013/14 onwards. The inclusion of two years of

historical data from 2010/11 means that the scenario outcomes differ between the two variants8.

7 Note that in the SNPP-2012 scenario, the historical MYEs are defined up to 2012; from 2012/13 onwards,

the population growth trajectory is as defined in the official projection. 8 If the results were presented from 2013 onwards, the dwelling-growth and household-growth outcomes

in the dwelling-led scenarios would be the same in the HH-12 and HH-08 variants.

26

July 2015

Northumberland: HH-12 Core Scenario Outcomes

Figure 12: Northumberland HH-12 core scenario outcomes: population growth 2001–2031

Table 4: Northumberland HH-12 core scenario outcomes (2011–2031)

Note: The forecast period is 2013/14–2030/31. Therefore, scenario outcomes presented for the 2011–2031 plan period include two years of historical data.

Scenario Population

Change

Population

Change %

Households

Change

Households

Change %

Net

MigrationDwellings Jobs

Jobs-led Policy 56,033 17.7% 32,126 23.2% 3,157 1,716 553

Jobs-led Upper 54,498 17.2% 31,507 22.7% 3,089 1,683 520

Jobs-led Baseline 43,121 13.6% 26,915 19.4% 2,585 1,438 275

Jobs-led Lower 36,835 11.6% 24,370 17.6% 2,307 1,302 140

PG-10yr 19,493 6.2% 16,733 12.1% 1,478 894 -237

Dwelling-led RSS 15,013 4.7% 15,613 11.3% 1,304 834 -342

PG-5yr 11,465 3.6% 13,363 9.6% 1,123 714 -427

SNPP-2012 8,097 2.6% 12,779 9.2% 970 683 -496

Dwelling-led Past Delivery 5,148 1.6% 11,609 8.4% 862 620 -555

Change 2011–2031 Average per year

27

July 2015

Northumberland: HH-08 Core Scenario Outcomes

Figure 13: Northumberland HH-08 core scenario outcomes: population growth 2001–2031

Table 5: Northumberland HH-08 core scenario outcomes (2011–2031)

Note: The forecast period is 2013/14–2030/31. Therefore, scenario outcomes presented for the 2011–2031 plan period include two years of historical data.

Scenario Population

Change

Population

Change %

Households

Change

Households

Change %

Net


Jobs-led Policy 56,033 17.7% 34,135 24.6% 3,157 1,823 553

Jobs-led Upper 54,498 17.2% 33,500 24.2% 3,089 1,790 520

Jobs-led Baseline 43,121 13.6% 28,790 20.8% 2,585 1,538 275

Jobs-led Lower 36,835 11.6% 26,178 18.9% 2,307 1,398 140

PG-10yr 19,493 6.2% 18,921 13.6% 1,478 1,011 -237

Dwelling-led RSS 11,646 3.7% 15,830 11.4% 1,156 846 -414

PG-5yr 11,465 3.6% 15,425 11.1% 1,123 824 -427

SNPP-2012 8,097 2.6% 14,413 10.4% 970 770 -496

Dwelling-led Past Delivery 2,005 0.6% 11,826 8.5% 724 632 -621


28

July 2015

Population growth under the core scenarios ranges from 1.6% and 0.6% under the Dwelling-led 4.5

Past Delivery scenario to 17.7% under the Jobs-led Policy scenario. These population growth

figures result in a dwelling growth range of between 620 and 1,716 dwellings per year under the

HH-12 alternative. Under the HH-08 outcomes, the average annual dwelling requirements are

higher, ranging from 632 to 1,823 dwellings per year.

The official benchmark scenario, the SNPP-2012 scenario, results in 2.6% population growth over 4.6

the 2011–2031 plan period. This results in an average annual dwelling requirement of 683 under

HH-12 and 770 under HH-08.

In the dwelling-led scenarios, the choice of headship rates affects the relationship between the 4.7

annual dwelling constraint and the resulting population growth. Population growth is lower

under the HH-08 outcome due to the headship rate trajectory resulting in a lower average

household size, i.e. the same number of dwellings is associated with a smaller population size.

With a lower annual average increase in the number of dwellings, the Dwelling-led Past Delivery

scenario results in lower population growth (0.6%–1.6% under the HH-08 and HH-12 alternatives

respectively) than the Dwelling-led RSS scenario (3.7%–4.7% HH-08 and HH-12 alternative

respectively).

The alternative trend scenarios result in population growth ranging from 3.8% under the PG-5yr 4.8

scenario to 6.2% under the PG-10yr scenario. The higher rate of growth under the PG-10yr

scenario is a reflection of the relevant historical time periods used in the calibration of future

migration assumptions: net internal migration was higher at the start of the 10-year historical

period than during the latter years (see Figure 3 on page 7). Under the PG-5yr scenario, this level

of population growth (3.6%) results in an average annual dwelling requirement of 714 per year

under the HH-12 alternative, and 824 per year under the HH-08 alternative. The higher expected

population growth under the PG-10yr scenario (6.2%) results in a dwelling requirement range of

between 894–1,011 dwellings per year (HH-12 and HH-08 respectively).

Population growth is highest under the jobs-led scenarios, ranging from 11.6% under the Jobs-led 4.9

Lower scenario, to 17.7% under the Jobs-led Policy scenario. This higher level of population

growth results in higher dwelling-growth outcomes: 1,302–1,716 per year under HH-12 and

1,398–1,823 under HH-08. In the jobs-led scenarios, the underlying employment assumptions

determine the size of the labour force and the proportion of the resident population in

employment. If there is insufficient population to meet the jobs-growth target, net internal

29

July 2015

migration is used to redress the balance. The annual average net migration under the jobs-led

scenarios ranges from 2,307 to 3,157 per year, compared to 970 per year under the benchmark

SNPP-2012 scenario.

Under all but the jobs-led scenarios, the average annual jobs-growth figures are negative, 4.10

indicating a reduction in the number of people of working age living in Northumberland (i.e.

reflecting the ‘ageing population’).

Jobs-led Sensitivity Scenario Outcomes

Of the nine core scenarios, population growth is highest under the jobs-led scenarios. In the jobs-4.11

led scenarios, the underlying economic assumptions (economic activity rates, commuting and

unemployment) determine the balance between jobs-growth and population growth. Four

sensitivities have been developed for three of the jobs-led scenarios (Baseline, Upper and

Policy), to evaluate the demographic implications of altering the unemployment rate and

commuting ratio assumptions:

SENS A: Commuting ratio reduces in line with the historical 2001–2011 trend.

SENS B: Commuting ratio reduces in line with the NOMIS line of best fit.

SENS C: Unemployment rate reduces to lowest historical value by 2031 and the

commuting ratio reduces in line with the historical 2001–2011 trend.

SENS D: Unemployment rate reduces to lowest historical value by 2031 and the

commuting ratio reduces in line with the NOMIS line of best fit.

For each of the three jobs-led scenarios, the sensitivity results are compared to each ‘core’ 4.12

alternative in the following tables:

Jobs-led Baseline (Table 6 for HH-12 and Table 7 for HH-08)

Jobs-led Upper (Table 8 for HH-12 and Table 9 for HH-08)

Jobs-led Policy (Table 10 for HH-12 and Table 11 for HH-08)

Of the four sensitivity scenarios, the SENS B alternative results in the highest level of population 4.13

growth. Reducing the commuting ratio over the forecast period reduces the net out-commute

from Northumberland, resulting in a lower level of population growth when compared to the

core scenario alternative. The SENS C alternative results in the lowest level of population growth.

In this sensitivity, the commuting ratio and unemployment rate reductions are greatest, reducing

30

July 2015

the scale of net internal migration to meet the defined jobs-growth target. This results in a lower

average annual dwelling requirement when compared to the relevant core scenarios and the

alternative jobs-led sensitivity scenarios.

It is important to note in the Jobs-led scenarios that the jobs growth targets are applied in each 4.14

year of the forecast period, from 2013/14 to 2030/31. These growth targets are taken directly

from the 'Headcount jobs' as defined in the St Chad's employment forecasts. The jobs growth

targets applied in each ‘core’ scenario alternative (i.e. Baseline, Upper, Policy) have been applied

in each of the associated sensitivity variants.

Over the forecast period (i.e. 2013/14–2030/31), population growth is determined by growth in 4.15

the number of jobs. Between 2011 and 2013, the population is defined by the mid-year

population estimates for Northumberland. In these historical years, the labour force and jobs

growth figures are derived from the population, using the economic assumptions applied (i.e. the

economic activity rates, unemployment rate and commuting ratio).

Whilst the population growth figures are the same historically, the different unemployment and 4.16

commuting ratio assumptions result in different levels of jobs growth over the 2011–2013 period.

When averaged over the plan period (i.e. 2011–2031), the average annual jobs growth figures

therefore differ between each ‘core’ Jobs-led scenario and the associated sensitivity variants.

Table 6: Northumberland Jobs-led Baseline HH-12 sensitivity scenario outcomes (2011–2031)

Scenarios are ranked in order of population growth

Population

Change

Population

Change %

Households

Change

Households

Change (%)

Net


Core 43,121 13.6% 26,915 19.4% 2,585 1,438 275

SENS B 23,412 7.4% 18,977 13.7% 1,697 1,014 320

SENS D 22,092 7.0% 18,478 13.3% 1,625 987 320

SENS A 9,067 2.9% 13,180 9.5% 1,049 704 353

SENS C 7,813 2.5% 12,706 9.2% 980 679 353

Jobs-led

Baseline

(HH-12)


31

July 2015

Table 7: Northumberland Jobs-led Baseline HH-08 sensitivity scenario outcomes (2011–2031)


Table 8: Northumberland Jobs-led Upper HH-12 sensitivity scenario outcomes (2011–2031)


Table 9: Northumberland Jobs-led Upper HH-08 sensitivity scenario outcomes (2011–2031)


Population

Change

Population

Change (%)

Households

Change

Households

Change (%)

Net


Core 43,121 13.6% 28,790 20.8% 2,585 1,538 275

SENS B 23,412 7.4% 20,675 14.9% 1,697 1,104 320

SENS D 22,092 7.0% 20,185 14.6% 1,625 1,078 320

SENS A 9,067 2.9% 14,756 10.6% 1,049 788 353

SENS C 7,813 2.5% 14,291 10.3% 980 763 353

Jobs-led

Baseline

(HH-08)


Population

Change

Population

Change %

Households

Change

Households

Change (%)

Net


Core 54,498 17.2% 31,507 22.7% 3,089 1,683 520

SENS B 34,011 10.8% 23,274 16.8% 2,167 1,243 565

SENS D 32,642 10.3% 22,757 16.4% 2,092 1,216 565

SENS A 19,108 6.0% 17,266 12.4% 1,494 922 598

SENS C 17,808 5.6% 16,774 12.1% 1,423 896 598

Change 2011–2031 Average per yearJobs-led

Upper

(HH-12)

Population

Change

Population

Change (%)

Households

Change

Households

Change (%)

Net


Core 54,498 17.2% 33,500 24.2% 3,089 1,790 520

SENS B 34,011 10.8% 25,079 18.1% 2,167 1,340 565

SENS D 32,642 10.3% 24,571 17.7% 2,092 1,313 565

SENS A 19,108 6.0% 18,942 13.7% 1,494 1,012 598

SENS C 17,808 5.6% 18,459 13.3% 1,423 986 598

Jobs-led

Upper

(HH-08)


32

July 2015

Table 10: Northumberland Jobs-led Policy HH-12 sensitivity scenario outcomes (2011–2031)


Table 11: Northumberland Jobs-led Policy HH-08 sensitivity scenario outcomes (2011–2031)


Migration Sensitivity Scenario Outcomes

The population, migration and job-growth outcomes of the PG-10yr SENS scenario are presented 4.17

in Table 12, with the PG-10yr scenario included for comparison. The increased level of internal in-

migration in the PG-10yr SENS scenario results in a higher level of population growth when

compared to the PG-10yr core scenario (7.3% compared to 6.2%). The working-age population is

still reducing in size over time (due to the ageing of the population), but the higher level of

internal in-migration means that the labour force reduces in size at a lower rate than that seen

under the core scenario alternative. This results in a lower annual average reduction in the

number of jobs (-157 compared to -237).

Population

Change

Population

Change (%)

Households

Change

Households

Change (%)

Net


Core 56,033 17.7% 32,126 23.2% 3,157 1,716 553

SENS B 35,441 11.2% 23,853 17.2% 2,230 1,274 598

SENS D 34,066 10.8% 23,333 16.8% 2,155 1,246 598

SENS A 20,463 6.5% 17,816 12.8% 1,555 952 631

SENS C 19,156 6.1% 17,322 12.5% 1,483 925 631

Jobs-led

Policy

(HH-12)


Population

Change

Population

Change (%)

Households

Change

Households

Change (%)

Net


Core 56,033 17.7% 34,135 24.6% 3,157 1,823 553

SENS B 35,441 11.2% 25,673 18.5% 2,230 1,371 598

SENS D 34,066 10.8% 25,162 18.1% 2,155 1,344 598

SENS A 20,463 6.5% 19,505 14.1% 1,555 1,042 631

SENS C 19,156 6.1% 19,020 13.7% 1,483 1,016 631

Jobs-led

Policy

(HH-08)


33

July 2015

Table 12: Northumberland PG-10yr SENS scenario outcome (2011–2031)

The higher level of population growth under the PG-10yr SENS scenario results in a higher annual 4.18

average dwelling requirement (Table 13), between 975 and 1,093 dwellings per year (HH-12 and

HH-08 outcomes respectively).

Table 13: Northumberland PG-10yr SENS scenario dwelling growth requirement (2011–2031)

Population

Change

Population

Change (%)Net Migration Jobs

Core: PG-10yr 19,493 6.2% 1,478 -237

PG-10yr SENS 23,225 7.3% 1,648 -157

Scenario

Change (2011–2031) Average per year

Household

Change

Household

Change %

Average

Annual

Dwellings

Household

Change

Household

Change %

Average

Annual

Dwellings

Core: PG-10yr 16,733 12.1% 894 18,921 13.6% 1,011

PG-10yr SENS 18,259 13.2% 975 20,469 14.8% 1,093

Scenario

HH-12 HH-08

34

July 2015

5 Summary

Requirements

Northumberland County Council has commissioned Edge Analytics to provide an updated range 5.1

of demographic scenarios for Northumberland, using the latest demographic and economic

assumptions.

Edge Analytics has produced a range of scenarios using POPGROUP v.4 technology. The 2012-5.2

based SNPP is included within this range as the official benchmark scenario.

Alternative trend-based scenarios have also been developed together with four jobs-led 5.3

scenarios in which population growth is determined by growth in the number of jobs defined by

the Policy Research Group (St Chad’s College Durham) forecasts. Two dwelling-led scenarios have

also been developed in which population growth is determined by the defined growth in the

number of dwellings, based on the RSS and the historical 5-years of dwelling delivery (2009/10–

2013/14).

Jobs-led sensitivity scenarios have been developed in which varying commuting ratio and 5.4

unemployment rate assumptions have been applied to assess the impact on population change

and associated dwelling growth. A migration sensitivity scenario has also been developed in

which the internal migration flows are adjusted to evaluate the impact of future increased levels

of migration to Northumberland from the surrounding area.

Scenario Summary

In all of the scenarios, household and dwelling growth has been assessed using assumptions from 5.5

both the 2012-based and 2008-based DCLG household projection models. In all scenarios, the

2012-based headship rates result in lower dwelling growth over the 2011–2031 plan period than

under the 2008-based headship rates.

35

July 2015

Using the 2012-based headship rates, the expected average annual dwelling requirement under 5.6

the core scenarios ranges from +620 under the Dwelling-led Past Delivery scenario to +1,716

under the Jobs-led Policy scenario (Table 14). Using the 2008-based headship rates, the dwelling

requirement ranges from 632–1,823 dwellings per year under the Dwelling-led Past Delivery and

Jobs-led Policy scenario respectively.

Table 14: Average annual dwelling requirement 2011–2031

Note: Sensitivity scenarios are shaded in blue. HH-08 refers to the dwelling growth outcomes using the 2008-based headship rates and HH-12 the 2012-based headship rates. Scenarios are ranked in order of dwelling growth under HH-12

HH-12 HH-08

Jobs-led Policy 1,716 1,823

Jobs-led Upper 1,683 1,790

Jobs-led Baseline 1,438 1,538

Jobs-led Lower 1,302 1,398

Jobs-led Policy SENS B 1,274 1,371

Jobs-led Policy SENS D 1,246 1,344

Jobs-led Upper SENS B 1,243 1,340

Jobs-led Upper SENS D 1,216 1,313

Jobs-led Baseline SENS B 1,014 1,104

Jobs-led Baseline SENS D 987 1,078

PG-10yr SENS 975 1,093

Jobs-led Policy SENS A 952 1,042

Jobs-led Policy SENS C 925 1,016

Jobs-led Upper SENS A 922 1,012

Jobs-led Upper SENS C 896 986

PG-10yr 894 1,011

Dwelling-led RSS 834 846

PG-5yr 714 824

Jobs-led Baseline SENS A 704 788

SNPP-2012 683 770

Jobs-led Baseline SENS C 679 763

Dwelling-led Past Delivery 620 632

Average Annual Dwelling Requirement (2011–2031)Scenario

36

July 2015

Issues for Consideration

The 2012-based DCLG household projection model has provided national and local authority 5.7

projections and assumptions for the total number of households by age-group and relationship-

status group (i.e. Stage One). DCLG intends to release additional data (Stage Two) which enables

disaggregation of these projections by each of seventeen household types, although a date for

the future release of this information has not been set. Whilst this new data will provide further

detail to the household outputs, it is not expected that they will change the household growth

assumptions implied by the Stage One output, which will continue to provide the controlling

totals for each local authority district.

It is recommended that the scenario outcomes are reconsidered when the Stage Two data is 5.8

released by DCLG, providing additional detail on the profile of growth by household-type implied

by the 2012-based household projection assumptions.

37

July 2015

Appendix A

POPGROUP Methodology

Forecasting Methodology

A.1 Evidence is often challenged on the basis of the appropriateness of the methodology that has

been employed to develop growth forecasts. The use of a recognised forecasting product which

incorporates an industry-standard methodology (a cohort component model) removes this

obstacle and enables a focus on assumptions and output, rather than methods.

A.2 Demographic forecasts have been developed using the POPGROUP suite of products. POPGROUP

is a family of demographic models that enables forecasts to be derived for population,

households and the labour force, for areas and social groups. The main POPGROUP model (Figure

14) is a cohort component model, which enables the development of population forecasts based

on births, deaths and migration inputs and assumptions.

A.3 The Derived Forecast (DF) model (Figure 15) sits alongside the population model, providing a

headship rate model for household projections and an economic activity rate model for labour-

force projections.

A.4 The latest development in the POPGROUP suite of demographic models is POPGROUP v.4, which

was released in January 2014. A number of changes have been made to the POPGROUP model to

improve its operation and to ensure greater consistency with ONS forecasting methods. The most

significant methodological change relates to the handling of internal migration in the POPGROUP

forecasting model. The level of internal in-migration to an area is now calculated as a rate of

migration relative to a defined ‘reference population’ (by default the UK population), rather than

as a rate of migration relative to the population of the area itself (as in POPGROUP v3.1). This

approach ensures a closer alignment with the ‘multi-regional’ approach to modelling migration

that is used by ONS.

A.5 For further information on POPGROUP, please refer to the Edge Analytics website:

http://www.edgeanalytics.co.uk/.

38

July 2015

Figure 14: POPGROUP population projection methodology

39

July 2015

Figure 15: Derived Forecast (DF) methodology

40

July 2015

Appendix B

Data Inputs & Assumptions

Introduction

B.1 Edge Analytics has developed a suite of demographic scenarios for Northumberland using

POPGROUP v.4 and the Derived Forecast model. The POPGROUP suite of demographic models

draw data from a number of sources, building an historical picture of population, households,

fertility, mortality and migration on which to base its scenario forecasts. Using historical data

evidence for 2001–2013, in conjunction with information from ONS sub-national population

projections (SNPPs) and DCLG household projections, a series of assumptions have been derived

which drive the scenario forecasts.

B.2 The following core scenarios have been produced:

SNPP-2012

PG-5yr

PG-10yr

Jobs-led Baseline

Jobs-led Lower

Jobs-led Upper

Jobs-led Policy

Dwelling-led RSS

Dwelling-led Past Delivery

B.3 The following jobs-led sensitivity scenarios have been produced:

Jobs-led Baseline SENS A

Jobs-led Baseline SENS B

Jobs-led Baseline SENS C

Jobs-led Baseline SENS D

Jobs-led Upper SENS A

Jobs-led Upper SENS B

Jobs-led Upper SENS C

Jobs-led Upper SENS D

Jobs-led Policy SENS A

Jobs-led Policy SENS B

Jobs-led Policy SENS C

Jobs-led Policy SENS D

41

July 2015

B.4 A migration sensitivity scenario (PG-10yr SENS) has also been produced.

B.5 In the following sections, a narrative on the data inputs and assumptions underpinning the

scenarios is presented.

Population, Births & Deaths

Population

B.6 In each scenario, historical population statistics are provided by the mid-year population

estimates (MYEs) for 2001–2013, with all data recorded by single-year of age and sex. These data

include the revised MYEs for 2002–2010, which were released by the ONS in May 2013. The

revised MYEs provide consistency in the measurement of the components of change (i.e. births,

deaths, internal migration and international migration) between the 2001 and 2011 Censuses.

B.7 In the SNPP-2012 scenario, future population counts are provided by single-year of age and sex

to ensure consistency with the trajectory of the ONS 2012-based SNPP.

Births & Fertility

B.8 In each scenario, historical mid-year to mid-year counts of births by sex from 2001/02 to 2012/13

have been sourced from ONS Vital Statistics.

B.9 In the SNPP-2012 scenario, future counts of births are specified to ensure consistency with the

official projection.

B.10 In the other scenarios, a ‘local’ (i.e. area-specific) age-specific fertility rate (ASFR) schedule, which

measures the expected fertility rates by age in 2013/14, is included in the POPGROUP model

assumptions. This is derived from the ONS 2012-based SNPP.

B.11 Long-term assumptions on changes in age-specific fertility rates are taken from the ONS 2012-

based SNPP.

42

July 2015

B.12 In combination with the ‘population-at-risk’ (i.e. all women between the ages of 15–49), the

area-specific ASFR and future fertility rate assumptions provide the basis for the calculation of

births in each year of the forecast period.

Deaths & Mortality

B.13 In each scenario, historical mid-year to mid-year counts of deaths by age and sex from 2001/02

to 2012/13 have been sourced from ONS Vital Statistics.

B.14 In the SNPP-2012 scenario, future counts of deaths are specified to ensure consistency with the

official projection.

B.15 In the other scenarios, a ‘local’ (i.e. area-specific) age-specific mortality rate (ASMR) schedule,

which measures the expected mortality rates by age and sex in 2013/14 is included in the

POPGROUP model assumptions. This is derived from the ONS 2012-based SNPP.

B.16 Long-term assumptions on changes in age-specific mortality rates are taken from the ONS 2012-

based SNPP.

B.17 In combination with the ‘population-at-risk’ (i.e. the total population), the area-specific ASMR

and future mortality rate assumptions provide the basis for the calculation of deaths in each year

of the forecast period.

Migration

Internal Migration

B.18 In all scenarios, historical mid-year to mid-year estimates of in- and out-migration by five year

age group and sex from 2001/02 to 2012/13 have been sourced from the ‘components of

population change’ files that underpin the ONS MYEs. These internal migration flows are

estimated using data from the Patient Register (PR), the National Health Service Central Register

(NHSCR) and Higher Education Statistics Agency (HESA).

B.19 In the SNPP-2012 scenario, future counts of internal migrants are specified, to ensure

consistency with the official projection.

43

July 2015

B.20 In the alternative trend scenarios, future internal migration flows are based on the area-specific

historical migration data. In the PG-5yr scenario, a five year internal migration history is used

(2008/09 to 2012/13). In the PG10yr scenario, a ten year history is used (2003/04 to 2012/13).

B.21 In the alternative trend scenarios (i.e. PG-5yr and PG-10yr), the relevant historical time period is

used to derive the age-specific migration rate (ASMigR) schedules, which are then used to

determine the future number of in- and out-migrants. In the case of internal in-migration, the

ASMigR schedules are applied to an external ‘reference’ population (i.e. the population ‘at-risk’

of migrating into the area). This is different to the other components (i.e. births, deaths, internal

out-migration), where the schedule of rates is applied to the area-specific population (i.e. the

population ‘at-risk’ of migrating out of the area). The reference population is defined by

considering the areas which have historically contributed the majority of migrants into the area.

In the case of Northumberland, it comprises all districts which cumulatively contributed 70% of

migrants into the area over the 2008/09–2012/13 period.

B.22 In the PG-10yr SENS scenario, internal in- and out-migrant counts are defined in each year of the

forecast period. The internal in-migration counts are based on the historical ten years of internal

migration with increases of +50 intervals per year between 2014/15 and 2020/21. From 2021/22

onwards the internal migration increases by +200 per year. The internal out-migration flows are

unaltered, remaining consistent with the PG-10yr scenario.

B.23 The Jobs-led and Dwelling-led scenarios calculate their own internal migration assumptions to

ensure an appropriate balance between the population and the targeted increase in the number

of jobs or dwellings that is defined in each year of the forecast period. A higher level of net

internal migration will occur if there is insufficient population and resident labour force to meet

the forecast number of jobs, or if there is insufficient population to meet the forecast number of

dwellings. In the Jobs-led and Dwelling-led scenarios, the profile of internal migrants is defined

by an ASMigR schedule, derived from the ONS 2012-based SNPP.

International Migration

B.24 Historical mid-year to mid-year counts of immigration and emigration by 5-year age group and

sex from 2001/02 to 2012/13 have been sourced from the ‘components of population change’

files that underpin the ONS MYEs. Any ‘adjustments’ made to the MYEs to account for asylum

cases are included in the international migration balance.

44

July 2015

B.25 Implied within the international migration component of change in all scenarios is an

'unattributable population change' (UPC) figure, which ONS identified within its latest mid-year

estimate revisions. The POPGROUP model has assigned the UPC to international migration as it is

the component with the greatest uncertainty associated with its estimation.

B.26 In all scenarios, future international migration assumptions are defined as ‘counts’ of migration.

In the SNPP-2012 scenario, the international in- and out-migration counts are drawn directly

from the ONS 2012-based SNPP.

B.27 In the alternative trend scenarios, the international in- and out-migration counts are derived

from the area-specific historical migration data. In the PG-5yr scenario, a five year international

migration history is used (2008/09 to 2012/13). In the PG-10yr and PG-10yr SENS scenarios, a ten

year history is used (2003/04 to 2012/13). In the PG-10yr and PG-10yr SENS scenarios, an

ASMigR schedule of rates is derived from a ten year migration history and is used to distribute

future counts by single year of age.

B.28 In the Jobs-led and Dwelling-led scenarios, international migration counts are taken from the

ONS 2012-based SNPP (i.e. counts are consistent with the SNPP-2012 scenario). An ASMigR

schedule of rates from the ONS 2012-based SNPP is used to distribute future counts by single

year of age.

Households & Dwellings

B.29 The 2011 Census defines a household as:

“one person living alone, or a group of people (not necessarily related) living at the

same address who share cooking facilities and share a living room or sitting room or

dining area.”

B.30 A dwelling is defined as a unit of accommodation which may comprise one or more household

spaces (a household space is the accommodation used or available for use by an individual

household).

B.31 Apart from in the Dwelling-led scenarios, the household and dwelling implications of the

population growth trajectory have been evaluated through the application of headship rate

45

July 2015

statistics, communal population statistics and a dwelling vacancy rate. These data assumptions

have been sourced from the 2001 and 2011 Censuses and the 2008-based and 2012-based

household projection models from the DCLG.

Household Headship Rates

B.32 A household headship rate (also known as household representative rate) is the “probability of

anyone in a particular demographic group being classified as being a household representative”9.

B.33 The household headship rates used in the POPGROUP modelling have been taken from the DCLG

2008-based and 2012-based household projections. The DCLG household projections are derived

through the application of projected household representative rates (also referred to as headship

rates) to a projection of the private household population.

B.34 In the scenarios presented here, headship rate assumptions have been sourced from the new

2012-based household projection model, and from the earlier 2008-based model, producing two

alternative outcomes for each scenario:

In the HH-12 outcome, the 2012-based DCLG headship rates are applied.

In the HH-08 outcome, the 2008-based DCLG headship rates are applied, scaled to be

consistent with the 2011 DCLG household total, but following the original trend

thereafter.

2012-based Headship Rates

B.35 The 2012-based headship rates have been sourced from the new 2012-based household

projection model from DCLG. The methodology used by DCLG in its household projection models

consists of two distinct stages:

Stage One produces the national and local authority projections for the total number

of households by sex, age-group and relationship-status group over the projection

period. All Stage One output and assumptions for the 2012-based household

projection model has been released by DCLG.

9 Household Projections 2012-based: Methodological Report. Department for Communities and Local Government (February 2015). https://www.gov.uk/government/statistics/2012-based-household-projections-methodology



46

July 2015

Stage Two provides the detailed ‘household-type’ projection by age-group, controlled

to the previous Stage One totals. Stage Two assumptions and output for the 2012-

based model have yet to be released by DCLG.

B.36 In POPGROUP, the 2012-based headship rates are defined by age, sex and relationship status.

These rates therefore determine the likelihood of person of a particular age-group, sex and

relationship status being head of a household in a particular year, given the age-sex structure of

the population.

2008-based Headship Rates

B.37 The 2008-based headship rates are provided by age-group and household type and therefore

define the likelihood of a particular household type being formed in a particular year, given the

age-sex profile of the population. Household-types are modelled with a 17-fold classification

(Table 15).

B.38 The 2008-based headship rates are scaled to the 2011 DCLG household total from the 2012-

based household projection model, following the original trend thereafter.

Table 15: Household type classification

ONS Code DF Label Household Type

OPM OPMAL One person households: Male

OPF OPFEM One person households: Female

OCZZP FAMC0 One family and no others: Couple: No dependent children

OC1P FAMC1 One family and no others: Couple: 1 dependent child

OC2P FAMC2 One family and no others: Couple: 2 dependent children

OC3P FAMC3 One family and no others: Couple: 3+ dependent children

OL1P FAML1 One family and no others: Lone parent: 1 dependent child

OL2P FAML2 One family and no others: Lone parent: 2 dependent children

OL3P FAML3 One family and no others: Lone parent: 3+ dependent children

MCZDP MIX C0 A couple and one or more other adults: No dependent children

MC1P MIX C1 A couple and one or more other adults: 1 dependent child

MC2P MIX C2 A couple and one or more other adults: 2 dependent children

MC3P MIX C3 A couple and one or more other adults: 3+ dependent children

ML1P MIX L1 A lone parent and one or more other adults: 1 dependent child

ML2P MIX L2 A lone parent and one or more other adults: 2 dependent children

ML3P MIX L3 A lone parent and one or more other adults: 3+ dependent children

OTAP OTHHH Other households

TOT TOTHH Total

47

July 2015

Communal Population Statistics

B.39 Household projections in POPGROUP exclude the population ‘not-in-households’ (i.e. the

communal/institutional population). These statistics are drawn from the DCLG 2012-based

household projections, which use statistics from the 2011 Census. Examples of communal

establishments include prisons, residential care homes and student halls of residence.

B.40 For ages 0–74, the number of people in each age group ‘not-in-households’ is kept fixed

throughout the forecast period. For ages 75–85+, the proportion of the population ‘not-in-

households’ is recorded (consistent with the DCLG approach). Therefore, the population not-in-

households for ages 75–85+ varies across the forecast period depending on the size of the

population.

Vacancy Rate

B.41 The relationship between households and dwellings is modelled using a ‘vacancy rate’, sourced

from the 2011 Census. The vacancy rate is calculated using statistics on households (occupied,

second homes and vacant) and dwellings (shared and unshared).

B.42 A vacancy rate of 6.4% for Northumberland has been applied, fixed throughout the forecast

period. Using this vacancy rates, the ‘dwelling requirement’ of each household growth trajectory

has been evaluated. This vacancy rate is slightly higher than the 2001 Census rate of 5.3%.

Labour Force & Jobs

B.43 Apart from in the Jobs-led scenarios, the labour force and jobs implications of the population

growth trajectory are evaluated through the application of three key data items: economic

activity rates, an unemployment rate and a commuting ratio.

Economic Activity Rates

B.44 The level of labour force participation is recorded in the economic activity rates. Economic

activity rates by five year age group (ages 16-74) and sex have been derived from 2001 and 2011

Census statistics. The 2011 Census statistics include an open-ended 65+ age categorisation, so

48

July 2015

economic activity rates for the 65–69 and 70–74 age groups have been estimated using a

combination of Census 2011 tables, disaggregated using evidence from the 2001 Census.

B.45 Rates of economic activity increased for women in all age groups between the 2001 and 2011

Censuses and in the older age groups for men (Figure 16).

Figure 16: Northumberland Economic activity rates: 2001 and 2011 Census comparison (source: ONS)

B.46 In all scenarios, Edge Analytics has made changes to the age-sex specific economic activity rates

to take account of changes to the State Pension Age (SPA) and to accommodate potential

changes in economic participation which might result from an ageing but healthier population in

the older labour-force age-groups.

B.47 The SPA for women is increasing from 60 to 65 by 2018, bringing it in line with that for men.

Between December 2018 and April 2020, the SPA for both men and women will then rise to 66.

Under current legislation, the SPA will be increased to 67 between 2034 and 2036 and 68

between 2044 and 2046. It has been proposed that the rise in the SPA to 67 is brought forward to

2026–202810.

B.48 ONS published its last set of economic activity rate forecasts from a 2006 base. These

incorporated an increase in SPA for women to 65 by 2020 but this has since been altered to an

accelerated transition by 2018 plus a further extension to 66 by 2020. Over the 2011–2020

period, the ONS forecasts suggested that male economic activity rates would rise by 5.6% and

10 https://www.gov.uk/changes-state-pension

https://www.gov.uk/changes-state-pension

49

July 2015

11.9% in the 60-64 and 65-69 age groups respectively. Corresponding female rates would rise by

33.4% and 16.3% (Figure 17).

Figure 17: ONS Labour Force Projection 2006 – Economic Activity Rates 2011–2020. Source: ONS

B.49 To take account of planned changes to the SPA, the following modifications have been made to

the Edge Analytics economic activity rates:

Women aged 60–64: 40% increase from 2011 to 2020.

Women aged 65–69: 20% increase from 2011 to 2020.

Men aged 60–64: 5% increase from 2011 to 2020.

Men aged 65–69: 10% increase from 2011 to 2020

B.50 Note that the rates for women in the 60–64 age and 65–69 age-groups are higher than the

original ONS figures (Figure 18), accounting for the accelerated pace of change in the SPA and

likely increases in female economic activity rates (between the 2001 and 2011 Censuses, female

participation rates increased across almost all age groups – see Figure 16). No changes have been

applied to other age-groups. In addition, no changes have been applied to economic activity rates

beyond 2020. This is an appropriately prudent approach given the uncertainty associated with

forecasting future rates of economic participation.

50

July 2015

B.51 Given the accelerated pace of change in the female SPA and the clear trends for increased female

labour force participation across all age-groups in the last decade, these 2011–2020 rate

increases (Figure 18) would appear to be relatively conservative assumptions.

Figure 18: Edge Analytics economic activity rate profiles for Northumberland, 2011 and 2020

comparison.

Commuting Ratio

B.52 The commuting ratio, together with the unemployment rate, controls the balance between the

number of workers living in a district (i.e. the resident labour force) and the number of jobs

available in the district.

B.53 A commuting ratio greater than 1.00 indicates that the size of the resident workforce exceeds the

number of jobs available in the district, resulting in a net out-commute. A commuting ratio less

than 1.00 indicates that the number of jobs in the district exceeds the size of the labour force,

resulting in a net in-commute.

B.54 From the 2011 Census ‘Travel to Work’ statistics, published by ONS in July 2014, commuting

ratios have been derived for Northumberland. This is compared to the 2001 Census value in

Table 16.

B.55 In the core scenarios, the 2011 Census commuting ratio has been applied, fixed throughout the

forecast period. Predicting how future commuting balances may change is complicated;

therefore, two sensitivities have been developed to examine a reduced net out-commute.

51

July 2015

Table 16: Commuting Ratio Comparison

Note: 2001 data from Census Table T101 – UK Travel Flows; 2011 data from Census Table WU02UK - Location of usual residence and place of work by age.

B.56 In the Jobs-led SENS A and SENS C scenarios, the commuting ratio incrementally decreases from

the 2011 Census value of 1.18, to 1.03 by 2031. This commuting ratio adjustment is based on the

rate of decline in the net out-commute seen between the 2001 and 2011 censuses (an annual

reduction of 0.007).

B.57 In the Jobs-led SENS B and SENS D scenario, the commuting ratio incrementally reduces from

1.18 in 2011, to 1.09 by 2031 (Figure 19). This commuting ratio adjustment is in line with the

NOMIS line of best fit, as defined by Northumberland County Council.

Figure 19: Comparison of the commuting ratios in the core and jobs-led sensitivity scenarios

Unemployment Rate

B.58 The unemployment rate, together with the commuting ratio, controls the balance between the

size of the labour force and the number of jobs available within an area.

Northumberland UA 0 2001 Census 2011 Census

Workers a 134,899 146,901

Jobs b 107,994 124,957

Commuting Ratio a/b 1.25 1.18

52

July 2015

B.59 In the core scenarios, the 2014 unemployment rate (6.6%) has been applied and incrementally

reduced to the ‘pre-recession’ average (2004–2007) of 5.0% by 2020 (Table 17), to account for

recovery following the recession. The unemployment rate has been fixed thereafter.

Table 17: Historical unemployment rates 2004–2014

B.60 In the Jobs-led SENS C and Jobs-led SENS D scenarios, the unemployment rate has been

incrementally reduced to the lowest historical unemployment rate of 4.4% by 2031 (Figure 20).

This reduction in unemployment is intended to reflect NCC plans for growth.

Figure 20: Comparison of the unemployment rates in the core and jobs-led sensitivity

scenarios

Northumberland

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14 Average

(2004–2007)

Unemployment Rate

(%)4.5 4.4 5.9 5.0 5.9 7.1 7.4 7.8 7.7 7.1 6.6 5.0

edge document template - northumberland...1.3 since the september 2014 forecasts were produced, new...

Documents