December - The Northern Way: Transport Compact
Transcription
December - The Northern Way: Transport Compact
NORTHERN WAY Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Report December 2006 Prepared for: Prepared by: The Northern Way Steer Davies Gleave 28-32 Upper Ground London SE1 9PD +44 (0)20 7919 8500 www.steerdaviesgleave.com Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Contents 1. 2. 3. 4. 5. 6. Page INTRODUCTION 1 Background 1 This Report 1 THE DYNAMIC URBAN MODEL 2 Introduction 2 Scope of the South and West Yorkshire Model 2 RUNNING THE SWYDM & VALIDATION 6 Validation of the base year (2001) 6 Validation of simulated years 2001 to 2004 9 Conclusions 13 TRANSPORT SCENARIOS TESTED 14 Model Results 18 PRODUCTIVITY IMPACTS 29 Agglomeration effects from the tested interventions 29 Relocation of jobs to more productive locations 31 CONCLUSIONS 33 Results 33 Land 33 Productivity impacts 33 FIGURES Figure 2.1 South and West Case Study Model Zoning System 4 Figure 3.1 Goodness of fit for Observed vs Modelled TTW Trips (District to District) 7 Figure 3.2 Jobs, Workforce and Population for Leeds 12 Figure 3.3 Jobs, Workforce and Population for Sheffield 12 Figure 4.1 Test 1 Generalised Time Changes 15 Figure 4.2 Test 2 Generalised Time Changes 17 Figure 4.3 Generalised Time vs Probability of Travel 18 Figure 4.4 Test 1 Change in Travel to Work to Leeds by Origin District 21 Figure 4.5 Test 1 in Travel to Work to Sheffield by Origin District 22 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 1 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Figure 4.6 Test 2 Change in Travel to Work to Leeds by Origin District 24 Figure 4.7 Test 2 Change in Travel to Work to Sheffield by Origin District 25 Figure 4.8 Test 1&2 Change in Travel to Work to Leeds by Origin District 27 Figure 4.9 Test 1&2 Change in Travel to Work to Sheffield by Origin District 28 TABLES Table 3.1 Total Travel to Work Trips into Leeds and Sheffield by Mode for 2001 7 Table 3.2 Observed & modelled jobs by District for 2001 8 Table 3.3 Observed and modelled Jobs numbers for 2001 to 2004 (000’s) 9 Table 3.4 Observed and modelled population 2001 to 2004 (000’s) 10 Table 3.5 Observed & modelled workforce 2001 to 2004 (000’s) 11 Table 4.1 Results For Test 1 versus Do Nothing 20 Table 4.2 Results For: Test 2 versus Do Nothing 23 Table 4.3 Results: Test 1&2 Versus Do Nothing 26 Table 5.1 Agglomeration effects from Test 1 (for 2010, in 2003 £pa) 30 Table 5.2 Agglomeration effects from Test 2 (for 2010, in 2003 £pa) 31 Table 5.3 Productivity gain from job relocations (£mpa) 32 APPENDICES A FURTHER INFORMATION ON VALIDATION \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 2 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 1. INTRODUCTION Background 1.1 A model, the South and West Yorkshire Dynamic Model (SWYDM) has been developed for the Northern Way to aid the understanding of the way transport can improve connectivity within and across the regions of the North of England, and translate this into effects on businesses, jobs and patterns of travel to work. 1.2 The model is based on an application of the Dynamic Urban Model for South and West Yorkshire previously developed and applied as part of research carried out for DfT on Transport and Business Locations1. The coverage of the model is not as great as the entire area encompassed by the Northern Way; it extends to South and West Yorkshire, with Manchester, Humberside and Merseyside as buffer areas. It does not include Tyne and Wear or Teesside at present, but could be further extended to cover this area as part of a second phase of work. This Report 1.3 The purpose of this report is to summarise the work undertaken using this model thus far and present the results of some strategic transport scenarios. 1.4 The rest of this report is structured as follows: • • • • • 1 Chapter 2 provides a more detailed overview of the structure of the Dynamic Urban Model and the scope of the current South and West Yorkshire Case Study model. Chapter 3 provides further detail relating to the model structure and the underlying assumptions within it. Chapter 4 provides the results of some strategic transport tests undertaken using the model to look, at a high level, at the impact of changes in transport costs brought about by investment, particularly how this could impact on the labour market and businesses. Chapter 5 extends the analysis to estimate the agglomeration and productivity benefits that derive from these changes. Chapter 6 summarises the key findings from the work undertaken thus far. The Impact of Transport on Business Location Decisions,2006 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 1 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 2. THE DYNAMIC URBAN MODEL Introduction 2.1 The Dynamic Urban Model has been developed over several years and used in a number of applications across the UK including the DfT’s South and West Yorkshire Case Study Model and the Merseyside Strategic Model developed to support the Merseyside Local Transport Plan in 2005/2006. 2.2 The model combines conventional transport modelling techniques with models of business formation, population migration and land-use, to provide a view on how all of these factors work together and interact over time. Starting from a base year of 20012, it moves through simulated time in small steps, calculating how events are likely to change in each step. The model provides the means to model processes that are much harder to handle in traditional equilibrium models, particularly where feedback effects are concerned, such as the effect of travel conditions on the attractiveness of a place to live. 2.3 The model is built using a general simulation package called Vensim, produced by Ventana Systems in the US3. The model is zonal, dividing the study area into zones of varying size. In each zone it keeps track of the population, the businesses, the infrastructure and the land. 2.4 The main dynamics in the model are created via the idea of the attractiveness of each zone for different types of activity. The model monitors each zone through time, and considers how attractive it is from four points of view: • • • • From the point of view of businesses and employers; From the point of view of households; From the point of view of developers, who build offices and other business premises; From the point of view of house builders, who provide the housing. 2.5 In general, the more attractive a zone is for an activity, the more that activity will take place, until new constraints, such as land, or the workforce, limit further growth. 2.6 A technical note is available that describes in further detail the mechanics and structures within the model4. Scope of the South and West Yorkshire Model 2.7 The model developed for the earlier DfT research project was not immediately suitable for the Northern Way study, and several extensions and enhancements had to be made. These included: 2 2001 tends to be used as the base year in most applications because of the wealth of information available from the Census to populate the model. 3 see www.vensim.com 4 The Dynamic Urban Model V4.doc \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 2 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model • • 2.8 The re-incorporation of household migration dynamics, which had been excluded from the original application; Re-basing the model to 2001 and validating the travel to work matrices for that year. The original version of the model had a base year of 1991 to meet the requirements of the DfT study. A total of 104 zones is included in the new model across seventeen administrative districts: Barnsley, Doncaster, Rotherham, Sheffield, Bradford, Calderdale, Kirklees, Leeds, Wakefield, York, Bolsover, Chesterfield, High Peak, North East Derbyshire, Harrogate, Selby and Bassetlaw. Eight buffer zones have been identified giving a total of 112 modelled zones. Counties adjacent to the modelled area were selected as buffers: the rest of Derbyshire, rest of North Yorkshire, rest of Nottinghamshire, Humberside, Manchester, rest of the North West, Lincolnshire and the West Midlands. The map overleaf illustrates the modelled area and the zone structure (i.e. the 104 internal zones). \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 3 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 2.1 SOUTH AND WEST CASE STUDY MODEL ZONING SYSTEM \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 4 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 2.9 The SWYDM reads data describing current conditions in the South and West Yorkshire area from a variety of Excel spreadsheets and text files, while information about the transport networks have, as part of the DfT research project, been extracted from the South and West Yorkshire Multi-Modal Study (SWYMMS) model in the form of generalised time matrices. 2.10 Three modes are represented in the model: highways, public transport and slow modes. Highway and public transport generalised cost matrices were taken from the SWYMMS model. No distinction is made between public transport modes; the generalised costs are composites of all available public transport modes. 2.11 The public transport costs comprised walk time, in vehicle time, average wait, interchanges and fares. Highway costs were made up of in-vehicle time and operating costs, where operating cost was a function of distance and pence per kilometre. The generalised times for slow modes were inferred using Census Travel to Work (TTW) data and the highway and public transport generalised costs5. 2.12 The SWYDM provides many outputs at the zone level describing jobs, workforce, businesses, households, population etc. However the main outputs that can be used to feed into the wider benefit calculations are: • • • 5 Relocation of jobs between zones, including the impact on individual industrial sectors. Estimates of how many more people get into work. At present the workforce (people in work or available for work) is a fixed proportion of the population, but a varying fraction of the workforce is in work at any time. Estimates of how the mix of job types on offer and the numbers of people occupying those jobs changes. It tries to match workforce skills, by category, with job skills required by employers. If the mix of job types changes, these jobs can only be filled if the workforce skills are available, and these may come via bigger catchments, inward migration or training of the existing workforce to convert them from one skills group to another. Although the SWYMMS model has over 500 zones, the granularity of the zoning is quite different from the modelled area we focussed on. Some SWYMMS zones fell into several of our modelled areas whereas others spanned more than one. We therefore undertook a process of infilling the generalised cost matrices so that each OD pair had an associated cost. A piece of commercial software, Drivetime, was used to produce car times for all zone pairs in the model. These times were then substituted where there was no SWYMMS information. We used the observed mode shares from the Travel to Work Data (TTW) in the 2001 Census together with the known highway travel costs to infer public transport costs for those flows where we did not have information from the SWYMMS model. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 5 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 3. RUNNING THE SWYDM & VALIDATION 3.1 Although the original model had been tested and validated, the extensions to it needed for this work meant that these tests had to be repeated, in the following way. • • • 3.2 To begin, it was necessary to run simulations with data from the 2001 Census to establish a stable base year position. We then ran the model for the period 2001 to 2005 to confirm it was replicating observed trends sufficiently well. Finally we ran the model into the future to test the transport interventions against a ‘do-nothing’ base case. The section that follows gives the results of the first two model runs and the level of validation obtained. Two validation tasks have been carried out after aggregating the zone figures to Districts: • • The first was checking how well the model replicated known Travel to Work (TTW) patterns in aggregate and by mode for 2001; The second was how well the model reproduced observed numbers of businesses, jobs, workforce and population in 2001. Validation of the base year (2001) Travel to Work Patterns and Mode Shares for 2001 3.3 This section describes the validation tests undertaken for TTW forecasts for 2001. Figure 3.1 shows the goodness-of-fit for observed versus modelled TTW trips. Detailed matrices for total TTW trips are given in Appendix Figures A1.1 and A1.2. Each point on the graph represents a District. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 6 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 3.1 GOODNESS OF FIT FOR OBSERVED VS MODELLED TTW TRIPS (DISTRICT TO DISTRICT) 250 y = 0.9788x + 0.2435 R2 = 0.9877 Observed 000s 200 150 100 50 0 0 50 100 150 200 250 Modelled 000s 3.4 The fit is good, with the points clustered around a diagonal line with a slope of 1 – as it should be – and a high R2. 3.5 Table 3.1 below shows observed and modelled mode shares for travel to work into Leeds and Sheffield districts: a good fit is observed across all modes. TABLE 3.1 TOTAL TRAVEL TO WORK TRIPS INTO LEEDS AND SHEFFIELD BY MODE FOR 2001 Zone category Car Public Transport Slow TOTAL Leeds observed 69% (223,395) 20% (65,506) 11% (33,761) 100% (322,662) Leeds modelled 70% (228,468) 19% (61,265) 11% (35,791) 100% (325,524) Sheffield observed 67% (140,855) 22% (47,237) 11% (22,173) 100% (210,265) Sheffield modelled 69% (148,598) 20% (44,318 11% (22,967) 100% (215,883) Jobs in 2001 3.6 Modelled jobs have been validated against Labour Market Statistics (LMS) accessed via Nomis. Table 3.2 shows the results. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 7 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE 3.2 Zone-category Observed Jobs Modelled Jobs % discrepancy Leeds 385,992 337,944 -12.45% Wakefield 122,199 108,809 -10.96% Calderdale 78,793 58,934 -25.20% Bradford 193,447 158,383 -18.13% Kirklees 152,984 114,512 -25.15% Sheffield 231,568 197,111 -14.88% Rotherham 89,026 77,267 -13.21% Doncaster 69,498 81,243 16.90% Barnsley 69,498 58,419 -15.94% 1,393,005 1,192,622 -14.38% York 101,904 65,948 -35.28% Selby 27,553 20,383 -26.02% NE Derbyshire 27,803 17,876 -35.70% Chesterfield 49,009 38,450 -21.55% Bolsover 18,827 13,946 -25.93% Harrogate 65,065 54,309 -16.53% Bassetlaw 43,952 30,789 -29.95% 1,727,118 1,434,323 -16.95% SWY Total TOTAL 3.7 OBSERVED & MODELLED JOBS BY DISTRICT FOR 2001 The model under-predicts the number of jobs across all areas with the exception of Doncaster. This has occurred as a result of model stabilisation when run with full dynamics: the model has found a position of stability with fewer jobs than actually was the case. Time did not allow us the opportunity to correct this, but for the purposes of this exercise, which was to provide a preliminary assessment of the effects of broad strategic initiatives, the near-equilibrium position of the model provides an adequate starting point. While suitable for the scenario testing undertaken as part of this work, further refinement of this base may be desirable if the model were to be used for scheme assessment. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 8 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Validation of simulated years 2001 to 2004 3.8 This section shows the validation tests undertaken for forecasts of jobs, population and workforce for the period 2001-2004. Actual time series data has been obtained from the Labour Market Survey for the purpose of comparison. Tables A1.3-A1.5 in the appendix tabulate the observed and modelled numbers of jobs, workforce and population for the years 2001 to 2004. Table 3.3, Table 3.4 and Table 3.5 show the modelled and observed values and changes between 2001 and 2004. In general the fit is reasonable, with the model under predicting growth in jobs a little in the metropolitan areas of South and West Yorkshire, but being correct in the wider areas of the city regions, while population numbers changed only very slightly, with the model replicating this stability. Again, while suitable for the scenario testing undertaken in this work, further model refinement may be desirable for detailed scheme assessments. TABLE 3.3 OBSERVED AND MODELLED JOBS NUMBERS FOR 2001 TO 2004 (000’S) Obs Jobs 2001 Obs Jobs 2004 % change Modelled Jobs 2001 Modelled Jobs 2004 % change Leeds 386 420 8.8 338 360 6.6 Wakefield 122 138 13.0 109 108 -0.4 District Calderdale 79 83 5.9 59 57 -3.9 Bradford 193 197 1.7 158 161 1.9 Kirklees 153 161 5.2 115 110 -3.6 Sheffield 232 246 6.3 197 195 -0.9 Rotherham 89 105 17.7 77 74 -3.8 Doncaster 69 75 8.3 81 85 4.5 Barnsley 69 75 8.3 58 56 -4.7 1,392 1,500 7.8 1,192 1,206 1.2 York 102 101 -1.2 66 65 -1.1 Selby 28 27 -0.4 20 19 -6.1 NE Derbyshire 28 25 -10.9 18 16 -9.7 Chesterfield 49 49 -0.8 38 38 -2.1 Bolsover 19 20 7.4 14 13 -5.3 Harrogate 65 65 0.4 54 54 -1.4 SWY Total Bassetlaw 44 41 -7.7 31 29 -6.4 Sub-total 335 328 -2.1 241 234 -2.9 1,727 1,828 5.8 1,433 1,440 0.0 TOTAL Note: Numbers may not sum due to rounding. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 9 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE 3.4 OBSERVED AND MODELLED POPULATION 2001 TO 2004 (000’S) Obs Pop 2001 Obs Pop 2004 % change Mod Pop 2001 Mod Pop 2004 % change Leeds 716 720 0.6 675 676 0.1 Wakefield 315 320 1.4 285 286 0.4 Calderdale 192 194 1.0 147 143 -2.5 Bradford 471 481 2.2 381 385 1.0 Kirklees 389 393 1.1 323 321 -0.6 Sheffield 513 516 0.6 473 473 -0.1 Rotherham 248 252 1.6 216 216 -0.2 District Doncaster 287 289 0.7 222 227 2.3 Barnsley 218 221 1.3 186 185 -0.5 3,349 3,386 1.1 2,908 2,912 0.1 York 181 185 2.0 155 155 0.0 Selby 77 77 0.8 47 44 -5.5 NE Derbyshire 97 97 0.5 57 51 -9.6 Chesterfield 99 100 0.9 93 94 0.1 Bolsover 72 73 1.9 47 44 -7.3 Harrogate 152 154 1.7 124 123 -0.7 Bassetlaw 108 110 2.3 72 67 -6.6 Sub-total 786 796 1.2 652 578 -11.4 4,135 4,182 1.1 3,503 3,490 -0.4 SWY Total TOTAL Note: Numbers may not sum due to rounding. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 10 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE 3.5 District OBSERVED & MODELLED WORKFORCE 2001 TO 2004 (000’S) Obs WF 2001 Obs WF 2004 % change Mod. WF 2001 Mod. WF 2004 % change Leeds 357 356 -0.4 278 278 0.1 Wakefield 147 155 5.5 117 117 0.4 Calderdale 95 92 -3.6 61 60 -2.6 Bradford 205 209 1.9 155 157 1.0 Kirklees 180 188 4.2 135 134 -0.6 Sheffield 237 235 -0.8 184 184 -0.1 Rotherham 117 118 0.8 87 87 -0.2 Doncaster 95 102 7.1 86 88 2.5 Barnsley 95 102 7.1 69 69 -0.6 1,528 1,557 1.9 1,172 1,174 0.2 York 94 92 -1.9 70 70 0.0 Selby 40 38 -5.0 21 20 -5.6 NE Derbyshire 45 46 2.9 24 22 -8.8 Chesterfield 45 46 1.8 38 38 0.1 Bolsover 32 31 -2.8 18 16 -7.5 Harrogate 76 77 1.7 55 55 -0.7 SWY Total Bassetlaw 48 52 8.1 31 28 -6.8 Outer Area T t l TOTAL 380 382 0.5 257 249 -3.2 1,908 1,938 1.6 1,429 1,423 -0.4 Note: Numbers may not sum due to rounding. 3.9 Figures 3.2 and 3.3, below, plot the observed and modelled trends between 2001 and 2004 for Leeds and Sheffield. Modelled behaviour is always below the actual, but the graphs show that the model is stable and producing gradual changes in jobs and population that are consistent with actual events. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 11 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 3.2 JOBS, WORKFORCE AND POPULATION FOR LEEDS Jobs, Workforce and Population for Leeds: Observed Vs. Modelled 800 Observed values in 000s 700 600 Modelled Jobs 500 Observed Jobs Modelled workforce 400 Observed Workforce Observed Population 300 Modelled population 200 100 0 2001 FIGURE 3.3 2002 2003 2004 JOBS, WORKFORCE AND POPULATION FOR SHEFFIELDF Jobs, Workforce and Population for Sheffield: Observed Vs. Modelled 600 Modelled Jobs 500 Observed Jobs 000s 400 Modelled workforce Observed Workforce 300 Observed Population Modelled population 200 100 0 2001 2002 2003 2004 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 12 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Conclusions 3.10 The model allows the population, workforce, jobs and travel to work patterns all to vary through simulated time as conditions change. The tests show that it replicates travel to work patterns well, and that although it under-estimates the actual numbers of jobs somewhat, it generates stable behaviour, with trends in jobs and population over 20012004 that closely match actual events. We conclude that it does provide a sound basis for preliminary testing of strategic policy directions. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 13 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 4. TRANSPORT SCENARIOS TESTED 4.1 Two transport scenarios have been tested in the SWYDM with a focus on the area extending from Leeds in the North to Sheffield in the South, including Rotherham, Leeds, Bradford, Wakefield and Sheffield. • • The first was to simulate interventions that reduced public transport and highway generalised times for trips within each of the Leeds and Sheffield City Regions; The other looked at reducing highway and public transport generalised times for trips between the two City Regions. Test 1: Urban Intervention 4.2 The first test reduces transport costs within each urban area. Generalised time reductions of 10% have been made uniformly between zones which form a contiguous urban area, around either Leeds or Sheffield. The coverage of this change is shown in the figure below; zones affected are shaded in green. This means that the generalised time of all travel between and within zones in each of the contiguous areas is reduced by 10%, but there is no change between the two separate areas. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 14 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.1 TEST 1 GENERALISED TIME CHANGES \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 15 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 2: Inter Urban Intervention 4.3 The second test is an inter-urban scheme with three levels of generalised time reduction. • • • 4.4 The biggest improvement is city centre to city centre, with a reduction of 15% between the zones shaded dark green in Figure 4.2. The cut is applied to road and rail travel between the two city centres. The next level is urban area to city centre and vice versa, with a 10% reduction. This is applied to movements between zones shaded light green and zones shaded dark green in Figure 4.2, except where they are in the same City Region; in other words, it applies to the Leeds urban area to Sheffield city centre or the Sheffield urban area to Leeds city centre. It is assumed that such trips benefit from generalised cost reductions on the highway network and city centre to city centre rail travel, but the total reduction is rather less than for city centre to centre trips because of the travel in the urban area. The third level of improvement is a 5% reduction for urban area to urban area, applied to trips between the two areas shaded light green. Such trips benefit from reduced inter-urban highway and public transport costs but experience either access/egress costs on the local highway network or PT feeder trips. In this test there are no changes to intra-urban area generalised times. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 16 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.2 TEST 2 GENERALISED TIME CHANGES \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 17 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Model Results 4.5 Results of these model runs now follow. In both cases the test is compared to the do nothing position with no changes to generalised time. The impacts on travel-to-work patterns and changes in employment are each reported. The results correspond to the point ten years after the change in generalised time is introduced. 4.6 Some assumptions have to be made about land availability. If jobs and/or the population are to grow, they need to be accommodated in business premises and housings units and these require land. The model is given estimates of the land occupied by buildings in the base year, but to generate forecasts we have to say how much additional land will be available for new developments. For this work we have assumed a 10% margin in each zone. Despite this, the availability of housing in some zones is constrained in future simulated years and some oscillatory behaviour occurs as demand exceeds supply and then falls back as the model compensates by reducing inward migration rates. To develop more accurate forecasts, additional work would be required to determine how much land might be available in each zone in future. This could form part of the second phase of model development. However, to indicate the likely form of the consequences of strategic changes to transport, which is the aim of this work, the current assumptions are a reasonable starting point. 4.7 Many of the responses to changes in generalised times are modelled using deterrence curves that plot how people’s willingness to accept travel costs falls as those costs rise. These curves are used for travel-to-work trips and business-to-business attractiveness. The travel-to-work deterrence function contained in the model has been calibrated for the local area and looks like the figure below where the x-axis indicates generalised time (GT) and the y-axis is the proportion of people willing to accept that GT for a commute trip. It can be seen that the impact of a given change in the generalised time depends on whether this occurs on a steep or a flatter part of the curve. 4.8 By and large we might expect Test 1 to focus more on shorter trips, of which a proportion will lie on the steeper part of the curve, where the marginal impact is greatest. Test 2, by contrast, focuses on trips more to the right hand end of the curve where the impact is less. FIGURE 4.3 GENERALISED TIME VS PROBABILITY OF TRAVEL \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 18 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 1: Urban Highway and PT Interventions 4.9 Changes in travel to work movements are greater for Leeds than for Sheffield, as Leeds is a stronger economic centre to begin with. The headline changes are: • • • • • • • 4.10 6 An increase in inter-district movements, in particular for Bradford and Wakefield, and corresponding decreases in intra-district movements with the largest changes for Wakefield, Kirklees and Leeds. Trips starting in Bradford and ending in Leeds increase by 3,055. Trips starting in Wakefield and ending in Leeds increase by 2,265. The net increase for trips into Leeds is close to 5,500 (at the bottom of the table) whereas for Sheffield trips fall by 750. Jobs are attracted into Leeds, which sees a growth of nearly 6,000, but many of these are transferred from the surrounding areas as activity becomes focussed in the city centre. The net increase in South and West Yorkshire jobs is more modest, at around 1,167. On the other hand the number of South and West Yorkshire residents in work rises by 1,923 as access to jobs improves. This has been achieved by reducing the pool of job vacancies. There is some growth in households across the region as it becomes a more attractive place to live, with increases in Bradford, Sheffield and Wakefield. Table 4.1 below shows the results of the urban intervention test. The last three columns indicate the change in jobs, businesses and households as a result of the intervention. The other six columns show the change in travel to work movements; for example the change in trips starting in Leeds and staying within Leeds is -1,325. The top half of the table shows the results for the core area, where the full model dynamics were allowed to operate, while the bottom half shows the results for the outer and buffer zones where jobs and population were held constant but travel to work patterns were allowed to vary6. The reason for this is to avoid model boundary effects. Zones close to the edge of the model have only part of their catchments represented, so if employers there find it harder to recruit from the core area, say, they will suffer more in the model than they might in reality because in practice they could seek to recruit from zones excluded from the model. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 19 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE 4.1 RESULTS FOR TEST 1 VERSUS DO NOTHING Intra District Trips Origin InterDistrict Trips Buffer Trips TOTAL Trips TTW Trips to Sheffield TTW trips to Leeds Jobs Busines s H Holds Barnsley 48 -34 -6 7 -171 53 -183 -15 8 Doncaster 70 -56 0 14 -51 16 -25 -35 6 Rotherham -346 305 0 -41 174 6 -132 -8 4 Sheffield -659 808 -9 140 -659 9 -975 -6 114 Bradford -977 2,108 0 1,130 -1 3,055 1,916 66 495 Calderdale -220 296 0 76 -1 69 -1,546 -111 16 Kirklees -1,361 1,429 0 67 -17 788 -2,463 -138 -4 Leeds -1,325 1,601 -2 273 -11 -1,325 5,983 391 -223 Wakefield -1,851 2,148 -42 256 -40 2,265 -1,408 -88 96 SWY Sub Total -6,623 8,606 -60 1,923 -778 4,936 1,167 55 513 Bassetlaw 18 -56 1 -37 -42 1 0 0 0 Bolsover 13 -48 1 -34 -53 0 0 0 0 Chesterfield 67 -98 2 -30 -127 0 0 0 0 High Peak 5 -5 0 0 -8 1 0 0 0 NE Derbyshire -238 283 -1 45 345 0 0 0 0 Harrogate -258 295 -4 32 0 276 0 0 0 Selby 18 -5 0 13 0 72 0 0 0 York -70 76 0 6 0 57 0 0 0 Gr. Manchester -131 239 0 108 -27 57 0 0 0 Buffer Zones -6 -11 -6 -23 -64 41 0 0 0 -583 670 -8 79 25 506 0 0 0 -7,206 9,276 -68 2,002 -753 5,442 1,167 55 513 Other Sub Total TOTAL 4.11 The maps below show the changes in travel to work movements into Leeds and Sheffield. The net change in trips into each district is shown in the text box in the top left hand corner of each diagram. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 20 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.4 TEST 1 CHANGE IN TRAVEL TO WORK TO LEEDS BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 21 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.5 TEST 1 IN TRAVEL TO WORK TO SHEFFIELD BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 22 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 2: Inter Urban Highway and PT Interventions between Leeds and Sheffield Regions 4.12 Table 4.2 the results of the inter-urban intervention test. As before, the last three columns indicate the changes in jobs, businesses and households while the other six columns show the change in travel to work movements. For example, the change in trips starting in Leeds and staying within Leeds is -28. 4.13 The scale of change is smaller than for test 1, but we see: • • • A modest increase in trips between districts and a slight reduction in trips within them; A modest increase in the number of people in work; and A small net increase in the number of jobs, with some displacement towards Rotherham and Sheffield, which gain, from the other Districts, such as Leeds and Kirless, which see a small reduction. TABLE 4.2 Origin RESULTS FOR: TEST 2 VERSUS DO NOTHING Intra District Trips InterDistrict Trips Buffer Trips TOTAL Trips TTW trips to Sheffield TTW trips to Leeds Jobs H. Holds Business 6 19 -2 24 34 -6 -10 0 8 Doncaster -33 39 0 6 10 0 -116 -6 1 Rotherham 82 -8 0 74 66 14 277 28 11 Sheffield -133 291 -7 152 -133 161 413 29 -16 Bradford -23 12 0 -11 8 9 -83 -5 -2 Calderdale -8 3 0 -4 0 3 -42 -4 -1 Kirklees -39 38 0 -1 9 7 -138 -7 0 Leeds -28 24 0 -4 51 -28 -199 -4 -2 Wakefield -75 82 -2 5 105 -21 27 8 1 SWY Sub Total -250 501 -11 240 150 140 130 38 1 Bassetlaw 1 26 -1 26 23 0 0 0 0 Bolsover -2 26 -1 23 22 0 0 0 0 Chesterfield 2 17 -1 18 55 0 0 0 0 High Peak 0 5 0 5 5 0 0 0 0 -73 106 0 32 87 3 0 0 0 Harrogate 0 -1 0 0 0 0 0 0 0 Selby 1 -1 0 0 0 0 0 0 0 York 0 0 0 0 0 0 0 0 0 Gr. Manchester -2 12 0 11 16 1 0 0 0 Buffer Zones 18 66 18 101 26 0 0 0 0 Other Sub Total -55 257 15 216 234 4 0 0 0 TOTAL -306 758 4 456 384 143 130 38 1 Barnsley NE Derbyshire \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 23 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.6 TEST 2 CHANGE IN TRAVEL TO WORK TO LEEDS BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 24 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.7 TEST 2 CHANGE IN TRAVEL TO WORK TO SHEFFIELD BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 25 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 1+Test2: Inter Urban and Within Region Improvements 4.14 As a final test we have combined both interventions. The result of this test is shown below. • • • The biggest gain in jobs is again in Leeds, although slightly less than for test 1, at 5,700. The net increase in jobs across SWY is slightly more than for test 1, while 2,500 more people are in work across the whole area. Outside the core SWY area we have again held the jobs and businesses fixed, as in the other tests, in order to avoid overstating model boundary effects. TABLE 4.3 Origin RESULTS: TEST 1&2 VERSUS DO NOTHING Intra District Trips InterDistrict Trips Buffer Trips TOTAL Trips TTW trips to Sheffield TTW trips to Leeds Jobs H. Holds Business Barnsley 53 -12 -8 33 -133 44 -199 -16 17 Doncaster 31 -10 0 21 -40 16 -150 -43 7 Rotherham -240 286 0 45 250 19 251 28 18 Sheffield -663 962 -16 283 -663 168 -473 28 90 Bradford -987 2,115 0 1,128 5 3,059 1,841 63 493 Calderdale -225 297 0 72 0 70 -1,585 -115 16 Kirklees -1,391 1,456 0 64 -8 788 -2,591 -146 -4 Leeds -1,366 1,633 -2 264 29 -1,366 5,742 383 -224 Wakefield -1,903 2,206 -43 260 48 2,232 -1,477 -92 96 SWY Sub Total -6,691 8,932 -70 2,171 -513 5,029 1,360 90 507 York 18 -26 0 -8 -16 1 0 0 0 Bolsover 11 -17 0 -6 -30 0 0 0 0 Chesterfield 45 -51 0 -7 -79 0 0 0 0 High Peak 4 2 0 6 -3 1 0 0 0 NE Derbyshire -238 331 -1 91 390 3 0 0 0 Harrogate -256 293 -4 32 0 273 0 0 0 Selby 19 -7 0 13 0 71 0 0 0 Bassetlaw -69 76 0 6 0 57 0 0 0 Gr. Manchester -134 253 0 119 -9 56 0 0 0 2 101 2 104 -35 41 0 0 0 -599 954 -5 350 218 503 0 0 0 -7,290 9,886 -75 2,521 -295 5,532 1,360 90 507 Buffer Zones Other Sub Total TOTAL \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 26 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.8 TEST 1&2 CHANGE IN TRAVEL TO WORK TO LEEDS BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 27 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model FIGURE 4.9 TEST 1&2 CHANGE IN TRAVEL TO WORK TO SHEFFIELD BY ORIGIN DISTRICT \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 28 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 5. PRODUCTIVITY IMPACTS 5.1 The analysis so far has described how changes in accessibility and attractiveness affect the location of jobs. But these changes have other important impacts on the economy. • Agglomeration economies. It is well documented that firms in denser locations are more productive. By locating in proximity, firms are more likely to interact, which aids the spreading of knowledge and technology. By locating where access to labour is better, firms are able to fill vacancies more quickly and they are more likely to find workers with a better skills match to what they are seeking. The effect density has on productivity is often referred to as agglomeration economies. Transport affects agglomeration through two routes. Transport improvements mean firms and workers are effectively located closer together, and the findings presented earlier in this report illustrate that jobs will relocate as a consequence of changes to transport – which will also affect the job density across locations. Both effects lead to productivity gains that are not recognised in standard transport appraisal. • Relocation of jobs to more productive locations. For various reasons, the productivity of similar types of activities differs across space (one of these being agglomeration economies). Typically jobs located in city centres tend to yield a higher output per worker than jobs elsewhere, even after correcting for differences in firms functions or workers’ skills. If commuting was not costly, workers would compete for jobs in these highly productive (and therefore highly paid) locations. However, commuting costs act as a constraint on the supply of labour to city centres. Where transport improvements enable the growth of employment in such highly productive locations, there are therefore additional productivity gains. 5.2 Guidance from the Department for Transport now allows us to quantify these productivity gains. We apply this guidance here to investigate how the two simulated transport interventions may affect the productivity of firms in the South and West Yorkshire area. Further detail about the agglomeration methodology can be found in a paper issued alongside this report.7 Agglomeration effects from the tested interventions 5.3 7 The interventions modelled reduce the cost of movement from and to selected zones within the modelled area. This, in itself, results in positive agglomeration effects for these locations. Secondly, the transport improvement causes jobs to relocate. This can cause further agglomeration effects, but can be either positive or negative for individual locations, depending on whether they are net gainers or losers of jobs. “Agglomeration Simulations of Northern Way Update”, SDG, December 2006 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 29 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 1 5.4 Table 5.1 summarises the impacts from Test 1 on the two City Regions. The figures are the net differences between the test case and the do-nothing. TABLE 5.1 AGGLOMERATION EFFECTS FROM TEST 1 (FOR 2010, IN 2003 £PA) Sheffield CR £pa Leeds CR £pa Barnsley 10,516 Harrogate 8,960 Doncaster -43,075 Selby -7,719 Rotherham 161,948 Leeds 8,494,366 Sheffield 1,694,806 Bradford 3,371,871 Bassetlaw -22,658 Calderdale -3,218 Bolsover -7,665 Kirklees 500,943 Chesterfield -15,380 Wakefield 1,819,598 NE Derbyshire 99,146 York -2,655 Barnsley 10,516 Total Productivity 1,877,638 Total Productivity 14,192,662 Per worker 2.9 Per worker 11.1 5.5 As predicted, the table shows that the areas directly affected by the transport improvement (primarily Sheffield and Leeds districts and some of the surrounding areas) see significant positive agglomeration effects. Other areas are negatively affected because they lose jobs to the central areas. The latter effect is to some extent exacerbated by the nature of the simulated intervention, for the zones that see transport cost improvements in this test are strictly delineated, being the green zones in Figure 4.1. In a real intervention we would expect zones outside the green areas also to gain some reductions in transport costs to the city centres and we might therefore still have found some positive agglomeration effects in the areas surrounding the two cities. 5.6 The most striking result of Test 1 is the distribution between the two City Regions. Despite the fact that the improvements affect the cost of movement in Leeds and Sheffield to a similar extent, only about 10% of the overall productivity gains from agglomeration fall to Sheffield. The contribution of the employment relocation from Sheffield to Leeds does not contribute materially to this finding. It appears that Leeds’s favourable position, both economically and geographically, means it can derive much larger agglomeration benefits from intra City Region improvements in transport. 5.7 As can be seen the overall effect is relatively small. Across the two City Regions, the per-worker increase in productivity is around £3pa for Sheffield and £11pa for Leeds. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 30 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Test 2 5.8 Table 5.2 presents the results from Test 2. Note that we do not have a benchmark measure of the relative size of the simulated interventions (such as scheme cost or the total potential benefits delivered). We can therefore not read anything into the relative total magnitude of the productivity impacts of the two tests. Our interpretation therefore focuses on the differences in the distribution of impacts across the city regions and districts. TABLE 5.2 AGGLOMERATION EFFECTS FROM TEST 2 (FOR 2010, IN 2003 £PA) Sheffield CR £pa Leeds CR £pa Barnsley -450 Harrogate -2,244 Doncaster -2,787 Selby -2,481 Rotherham 191,458 Leeds 2,031,580 Sheffield Bassetlaw 2,484,739 907 Bradford 689,683 Calderdale -4,787 Bolsover 2,232 Kirklees 99,015 Chesterfield 2,279 Wakefield 474,683 NE Derbyshire 85,704 York Barnsley Total Productivity Per worker 2,764,081 4.2 Total Productivity Per worker -8,394 -900 3,276,154 2.6 5.9 We see the same general pattern across the districts for Test 2 as we did for Test 1. The districts directly affected by the simulated changes in transport costs see significant positive productivity gains, whilst some of the peripheral areas are marginally worse off due to loss of employment to the centres. 5.10 Again, it is striking to see that the findings mirror those from the employment impacts. When transport is improved between the City Regions the economic impacts tend to be more equally distributed, with the net increases now being much more evenly balanced. Leeds still receives the largest share of the total productivity gains, but the impact per worker is larger in Sheffield. Relocation of jobs to more productive locations 5.11 8 The findings from the DUM model presented earlier in this report tell how jobs relocate within the South and West Yorkshire area following the two interventions. What we then need to know is how the productivity differs across locations. To be more specific, we need to know how much more or less productive a given worker will be in different locations. This is not an easy task. The only evidence that does exist8 has a relatively crude geographical disaggregation, only distinguishing between “Estimated Wage Relativities for England”, NERA, February 2002 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 31 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model South Yorkshire, West Yorkshire and Rest of Yorkshire. The evidence suggests that jobs located in West Yorkshire are about 7% more productive than equivalent jobs located in South Yorkshire and Rest of Yorkshire. 5.12 As a result, interventions that tend to lead to relocation of jobs to West Yorkshire will tend to lead to increased productivity overall, while attracting jobs away from West Yorkshire will reduce productivity. Table 5.3 illustrates what this means for the two interventions tested. TABLE 5.3 PRODUCTIVITY GAIN FROM JOB RELOCATIONS (£MPA) Test 1 Test 2 Leeds CR 51.1 -7.9 Sheffield CR -42.9 8.2 Other modelled areas -1.3 -0.4 Total 6.9 -0.1 5.13 The intra City Region test causes relocation of jobs from Sheffield and other modelled areas to Leeds and, since West Yorkshire as a location is more productive than other modelled locations, this leads to a net gain in productivity. The net gain is significant, but smaller than the gains from agglomeration. 5.14 The inter City Region test shows the opposite; a few jobs relocate from Leeds to Sheffield and other modelled areas, resulting in a net productivity loss, albeit insignificantly small. 5.15 These results suggest a potential conflict between overall growth in productivity and distributional impacts. While inter City Region improvements may lead to a more balanced growth across cities, there is the potential for larger productivity gains overall by encouraging the growth of cities that are already more productive. However, we would also repeat the health warning that comes with this calculation. The productivity per worker data on which they are based is at a scale much more aggregate than that which the model operates. There will therefore be productivity differences within West and South Yorkshire that are not adequately addressed in this calculation. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 32 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model 6. CONCLUSIONS Results 6.1 For the sub-region of South and West Yorkshire the first two tests produce quite different patterns of impact. Improving access to the city centres (test 1) has tended to concentrate activity in Leeds, driven largely by the improvement in commuting and thus the ability of centrally located employers to recruit. The converse is that it becomes rather harder for employers located outside central Leeds to recruit, and thus many of the ‘new’ jobs in Leeds are, in effect, transferred from the neighbouring areas. Sheffield was not able to capitalise on the changes in the same way, and while the reasons for this needs further investigation, it seems likely that they are due to the relative sizes of the two city’s economies in the base year. 6.2 The biggest change is the number of people in work, as indicated by the total number of TTW trips. This has been absorbed without an equal increase in job numbers by reducing the pool of job vacancies. 6.3 Test 2 has less impact on commuting because it affects longer trips (see Figure 4.3) and its consequences are more driven by how it affects business-to-business activities. Sheffield and Rotherham are now the main gainers, in terms of jobs, although the changes across the model are quite modest. 6.4 It appears that combining the two tests produces impacts that are largely the sum of the two9. This follows from the way the two tests have different impacts on businesses and employment, test 1 primarily affecting commuting, test 2 primarily affecting business to business activity. Land 6.5 The model is currently constrained by land availability, both for businesses and households and revising the assumptions used here may lead to different results. However we would argue that the types of responses reported here are realistic and credible and give a good indication of the ways different interventions can shape outcomes. In the past we have undertaken reviews of local plans and Regional Spatial Strategies (RSS) in order to model available land as realistically as possible. However this is a fairly significant task and thus was somewhat outside of the scope of what could be achieved in the time available for this work. This is a topic that might be revisited if the work were to be extended later. Productivity impacts 6.7 9 We have found that the conclusions relevant to the changes in employment largely hold true for productivity impacts from increased agglomeration. Whilst most areas are positively affected in both tests, the intra City Region improvements lead to gains to the stronger of the two economies, Leeds, that far outweighed those to Sheffield. Inter City Region improvements, however, seem to benefit both economies to a much Although not reported in Chapter 5, this applies to the productivity gains too. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 33 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model more similar extent. 6.8 On the other hand, jobs located in Leeds tend to be more productive than equivalent jobs in Sheffield. Since the intra City Region intervention tend to cause jobs to relocate to Leeds this means an overall productivity gain. The inter City Region test causes the opposite to happen – a net relocation of jobs from Leeds to Sheffield and a corresponding reduction of productivity. These finding suggests a possible conflict between distributional and overall productivity objectives. 6.9 However, the productivity impacts of job relocations is based on much cruder evidence than the agglomeration analysis and this conclusion should be treated with some caution. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc 34 Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model APPENDIX A FURTHER INFORMATION ON VALIDATION \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE A1.1 MODELLED 2001 DISTRICT-DISTRICT TRAVEL TO WORK TRIPS (EXCLUDING BUFFER ZONES) Sheffield Bradford Calderdale Kirklees Leeds Wakefield York Bolsover Chesterfield NE Derbyshire Harrogate Selby Bassetlaw Doncaster Rotherham Barnsley 42,090 3,389 5,443 5,971 362 46 2,782 3,567 9,272 19 36 24 38 6 107 101 1,503 87,377 4,871 1,334 108 6 122 611 2,413 105 43 30 20 12 713 969 Barnsley Zone Name Doncaster Destination Rotherham 5,251 8,539 44,393 23,440 37 3 162 254 759 5 1,445 413 328 1 38 4,123 Sheffield 3,436 1,259 12,363 157,536 57 7 276 388 635 2 1,629 2,844 4,666 1 11 1,018 Bradford Calderdale Kirklees Leeds Wakefield York Bolsover Chesterfield NE Derbyshire Harrogate Selby Bassetlaw 59 15 5 10 110,507 9,008 6,595 33,329 1,034 19 0 0 0 276 32 0 136 8 5 15 14,433 44,897 7,615 4,977 493 5 0 0 0 8 11 0 3,603 120 180 588 13,348 5,377 96,113 16,589 9,857 50 4 3 6 12 46 9 532 197 32 80 23,784 555 5,491 232,936 8,793 1,094 1 0 1 2,923 1,362 6 9,690 2,765 621 598 2,484 172 6,981 18,361 75,555 147 7 5 5 44 1,106 53 6 27 1 1 54 2 19 1,712 113 70,620 0 0 0 511 3,157 1 97 208 3,482 3,884 3 0 12 17 38 0 5,668 5,385 882 0 3 3,077 72 119 965 7,876 2 0 10 14 28 0 3,130 18,088 6,762 0 2 649 159 99 793 11,577 2 0 11 14 28 0 781 7,325 12,929 0 1 249 8 12 1 3 565 9 56 9,852 140 883 0 0 0 43,833 3,851 0 88 769 27 16 112 5 74 2,756 2,082 4,790 1 1 0 110 19,638 12 163 2,619 3,176 2,321 4 0 10 33 123 3 803 225 67 0 20 29,351 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE A1.2 OBSERVED 2001 DISTRICT-DISTRICT TRAVEL TO WORK TRIPS (EXCLUDING BUFFER ZONES) Bolsover Chesterfield NE Derbyshire Harrogate Selby Bassetlaw 150 434 66 155 131 147 1,236 1,715 88 313 528 565 72 90 2,067 872 150 378 2,822 2,885 54 81 882 20,682 1,459 281 3 0 6 807 138 3 4,295 762 84 3 6 0 63 36 6 102,168 16,865 5,868 206 21 27 21 185 127 27 5,891 230,826 8,840 1,775 21 27 21 3,693 1,658 42 613 5,346 20,644 83,081 382 36 45 30 325 2,659 116 103 149 4,429 466 58,887 3 12 0 1,526 1,618 9 52 33 12 8,631 2,858 1,274 9 6 2,076 102 48 9 1,164 24,856 3,842 9 9 344 88 15 1,361 8,528 11,472 9 15 356 427 1,578 3 3 3 44,696 346 9 2,721 4,045 6 6 0 519 14,995 22 172 27 380 299 147 12 48 27,534 6,803 420 193 1,770 3,083 6,043 6,066 3,697 214 108 229 2,048 2,525 55,796 22,245 143 70 323 1,097 741 9,863 158,541 265 113 543 2,211 91 64 222 126,898 4,496 3,889 42 45 117 6,885 53,297 6,019 1,503 211 289 901 8,560 8,063 580 444 339 763 14,882 2,009 2,489 1,114 518 831 1,763 York 42 210 21 92 253 Bolsover 39 83 500 1,204 12 12 27 Chesterfield 51 135 428 2,789 21 0 12 161 238 1,138 9,283 54 18 24 210 66 63 30 102 1,347 127 249 9,457 Selby 109 424 36 105 280 82 174 6,159 Bassetlaw 157 2,202 1,464 1,768 42 18 27 193 Doncaster 4,843 Barnsley Wakefield 374 Leeds 54 Kirklees 97 Calderdale 119 Bradford 30 Sheffield 81 Rotherham York Destination 50,640 2,159 Doncaster 1,774 77,214 Rotherham 2,922 4,759 Sheffield 3,165 1,928 Bradford 178 96 Zone Name Barnsley Calderdale Kirklees Leeds Wakefield NE Derbyshire Harrogate The fit for some districts may appear poor but it should be noted that the total movements are often small e.g. Bassetlaw to Bradford is only 42 trips. The Mean Square Error of observed versus modelled TTW trips is 73%. This appears high, but the variation in magnitude of TTW trips across destinations is very large. The model replicates the order of magnitude of TTW trips with reasonable accuracy. \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE A1.3 OBSERVED AND MODELLED JOBS FOR 2001 TO 2004 Observed jobs Modelled jobs Origin Zone 2001 2002 2003 2004 2001 2002 2003 2004 Leeds 386 400 405 420 338 343 351 360 Wakefield 122 125 130 138 109 108 108 108 Calderdale 79 79 81 83 59 58 57 57 Bradford 193 195 196 197 158 161 160 161 Kirklees 153 153 153 161 115 113 112 110 Sheffield 232 234 240 246 197 196 195 195 Rotherham 89 92 97 105 77 76 75 74 Doncaster 69 74 75 75 81 82 83 85 Barnsley 69 74 75 75 58 57 56 56 York 102 100 104 101 66 66 66 65 Selby 28 28 31 27 20 20 19 19 NE Derbyshire 28 26 26 25 18 17 17 16 Chesterfield 49 47 48 49 38 38 38 38 Bolsover 19 18 20 20 14 14 13 13 Harrogate 65 67 68 65 54 54 54 54 Bassetlaw 44 46 43 41 31 30 29 29 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE A1.4 OBSERVED AND MODELLED POPULATION FOR 2001 TO 2004 Observed population Modelled population Origin Zone 2001 2002 2003 2004 2001 2002 2003 2004 Leeds 716 716 715 720 675 675 675 676 Wakefield 315 317 318 320 285 285 286 286 Calderdale 192 193 193 194 147 145 144 143 Bradford 471 474 478 481 381 383 384 385 Kirklees 389 391 391 393 323 322 322 321 Sheffield 513 513 513 516 473 474 474 473 Rotherham 248 250 252 252 216 217 217 216 Doncaster 287 288 288 289 222 224 225 227 Barnsley 218 219 220 221 186 186 186 185 York 181 182 183 185 155 155 155 155 Selby 77 77 77 77 47 46 45 44 NE Derbyshire 97 97 97 97 57 55 53 51 Chesterfield 99 99 100 100 93 94 94 94 Bolsover 72 73 73 73 47 46 45 44 Harrogate 152 152 153 154 124 123 123 123 Bassetlaw 108 109 109 110 72 70 68 67 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model TABLE A1.5 OBSERVED AND MODELLED WORKFORCE FOR 2001 TO 2004 Observed workforce Modelled workforce Origin Zone 2001 2002 2003 2004 2001 2002 2003 2004 Leeds 357 348 349 356 278 278 278 278 Wakefield 147 147 156 155 117 117 117 117 Calderdale 95 95 92 92 61 61 60 60 Bradford 205 204 213 209 155 156 156 157 Kirklees 180 187 189 188 135 135 135 134 Sheffield 237 247 248 235 184 184 184 184 Rotherham 117 115 116 118 87 87 87 87 Doncaster 95 94 96 102 86 86 87 88 Barnsley 95 94 96 102 69 69 69 69 York 94 93 93 92 70 69 70 70 Selby 40 39 39 38 21 21 21 20 NE Derbyshire 45 47 51 46 24 23 23 22 Chesterfield 45 50 51 46 38 38 38 38 Bolsover 32 32 33 31 18 17 17 16 Harrogate 76 76 77 77 55 55 55 55 Bassetlaw 48 51 47 52 31 30 29 28 \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Appendix Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model CONTROL SHEET Project/Proposal Name: NORTHERN WAY Document Title: Model Development and Results for Northern Way using the South & West Yorkshire Dynamic Model Client Contract/Project Number: SDG Project/Proposal Number: 206745 ISSUE HISTORY Issue No. Date Details 1 18/12/2006 Final version of issue REVIEW Originator: Andrew Davies Other Contributors: Lars Rognlien, John Swanson, Neil Chadwick Review By: Print: John Swanson Sign: (electronic approval) DISTRIBUTION Clients: Steer Davies Gleave: \\Douglas\Work\projects\6700s\6745\Work\Model Development\01 SWYDM\02 Memo\M02 - SWYDM for Phase 1 Model Results V2 6.doc Control Sheet