The development scenario model (Appendix C)

Description of the model

The model, a Microsoft Excel^(r) spreadsheet, works in five steps, shown in Fig. C.1. Each is discussed in turn.

Fig. C.1: Operation of the model

Step 1: Allocate land to use categories

Land use categories

The gross land area is divided into a hierarchy of use categories. (See Fig C.2.)

Fig. C.2: Hierarchy of land use categories

Quantifying undevelopable land

The gross land area is separated into developable and undevelopable land. Undevelopable land includes designated natural heritage features and associated systems, as well as existing and proposed "fixed" infrastructure -- highway, rail, and utility corridors.

To provide a realistic sense of the impact of natural heritage features on density, six district-scale existing parcels of land in the Toronto metropolitan region were analyzed. Each is adjacent to existing urbanized areas or lies within or adjacent to areas designated for future urban development. To capture a range of different conditions with respect to natural heritage features, each is in a different part of the metropolitan region. These lands are likely to be developed in the near-to-medium term and are therefore likely to be subject to today's regulations and standards. Fig. C.3 shows the values for each area. The areas themselves are mapped in Fig. C.4.

Information on natural heritage features is drawn from the Neptis Foundation study, The State of Greenlands Protection in South Central Ontario (Fraser & Neary 2004), which defined "greenlands" as terrestrial and water-based features such as woodlands, wetlands, valleys, watercourses, and bodies of water, as well as conservation areas, agricultural preserves, or Crown land that are specifically designated by municipal, provincial, and federal governments and agencies (9-10).

Fig. C.3: Areas of pre-existing land uses in hectares and as % of gross land area

Designation does not equal protection. In areas under development pressure, designations can be, and are, changed (Fraser & Neary 2004:115). For each study area, the natural heritage features in the Neptis greenlands database were quantified regardless of their degree of real protection. These are considered the "core" of the natural heritage system.

Purpleville and Trafalgar have the largest amounts of natural heritage features and natural heritage systems as a proportion of gross land area -- approximately 27% and 39%, respectively. Mount Pleasant has the lowest amounts of each, 5% and 10%. Puslinch, North Brooklyn, and South Ancaster have the same proportion for natural heritage features (16%), though their natural heritage systems values differ, ranging from 24% to 35%.

The development scenarios are run on three hypothetical land bases derived from the six cases, each representing different levels of natural heritage features and natural heritage systems coverage: Low, Medium, and High.

Fig. C.4: Study area natural heritage system maps

Study area locations

Developable land

Once undevelopable land is excluded, the remainder is considered developable. Developable land is of two types: public and private property. Public property consists of local rights-of-way (including arterial roads but not limited-access expressways) and public facilities such as parks, community centres, and schools. The allocation of land for public facilities occurs in Step 3. Private property consists of residential and employment parcels, excluding associated rights-of-way.

Definition of employment land

It is important to note that the definition of employment land used here differs from that typically used in Ontario planning policy. Traditionally, Ontario planning policies define "employment land" as specialized, non-residential zones containing manufacturing, warehousing, and some types of commercial enterprises, but not retail. For the purposes of this model, "employment land" refers to any parcel that contains jobs in a single-use (as opposed to mixed-use) format. By this definition, an office building, shopping mall, or "big box" retail power centre qualifies as employment land, while a retail or office establishments within a residential building does not.

Treatment of vacant parcels

The model assumes that the gross and net densities include any vacant parcels. The net density of built parcels must be high enough to compensate for the depressing effect of vacant parcels on gross density.

Step 2: Quantify dwelling units and resident population

Dwelling units

The number of dwelling units that will fit into the residential lot area is derived from the "housing type mix" (the proportion of all dwelling units of each unit type) and the average land area per dwelling unit by type. The calculation has four steps:

1. The housing type mix is translated into an "area mix" by multiplying the housing type mix percentage for each type by the land area per unit for each type and dividing each result by the total area for all types. The resulting "area mix" is the proportion of residential land taken up by each type.

2. Multiplying the residential parcel area by the area mix produces the total land area occupied by each unit type.

3. Dividing the land area occupied by each unit type by the corresponding land area per unit for each type produces the number of units of each type.

4. The sum of these totals is the total number of residential units.

Resident population

The resident population is determined by multiplying the quantity of units of each housing unit type by the associated average household size.

Step 3: Optimize population and public facilities

Land for parks and schools is allocated in proportion to population or households. If the amount of land allocated to parks and schools is increased, and all other land allocations are held constant, the residential parcel area will decrease. As a result, the amount of land for public facilities and the size of the residential population must be brought into balance. This is done manually in the Excel spreadsheet, avoiding the potential complications of using auto-optimizing software utilities such as Solver. An auto-optimizing algorithm could, however, be incorporated into the model (see Ottensmann 2000).

To compare the impact of different input assumptions, the land area for parks and schools per quantity of residents or dwelling units is determined in several ways:

Pro rata calculation from the Central Pickering background studies. As part of the preparation of the Central Pickering Development Plan (MMAH 2006), consultants determined the land requirements for public facilities to serve a forecast population of 69,000 (MMAH 2005d).

2. Schools standards. Land area per school and number of schools required per 1,000 dwellings for elementary and secondary public and Catholic schools are cited in municipal planning documents.

3. Statutory conveyance standards for parks. Section 42 of the Ontario Planning Act specifies a standard of 5% of "neighbourhood land" (i.e., the developable area exclusive of employment lands) plus 2% of employment lands, or 1 hectare per 300 units to be set aside for parks.

4. Official Plan parks standards. Most municipal official plans set standards for a hierarchy of parks, each with a minimum land area and catchment area. These were compared to the Central Pickering and Planning Act standards, but due to wide differences among municipal formulas, these were not used in the model.

Combining the calculated school and park allocations produces two totals:

1. Pro rata amount (Central Pickering)

2. Planning Act parks standards + schools standards

The total land area required for public facilities is then expressed as a percentage of the developable land area. The larger of the two values is then used in the model.

Step 4: Quantify employment by type and location

Within the model, there are three types of jobs: those located in the home, those located on segregated employment lands, and the remainder, which are located in mixed-use settings. The proportions of jobs located in mixed-use settings and on employment lands is determined on the basis of documentary research (see Appendix C.2).

1. Jobs located in the home. Jobs located in the home are calculated in proportion to the employed labour force.

2. Jobs located on segregated employment land. The number of jobs on employment land is calculated in a manner similar to the way dwelling units are allocated on the residential lot area. A "job mix" -- the proportion of jobs in major office, business parks, and retail -- is translated into an "area mix," using known employment densities. The area mix is then used to determine the amount of employment land occupied by each job type. The total number of jobs by type is determined by multiplying the job densities of each employment area type by the amount of land used by each type. Research shows that employment lands are rarely 100% occupied. In projections, employment lands are generally assumed to have a "natural" rate of vacancy in order to maintain a fluid land market. An Oakville Economic Development Alliance report (2000:10) states that a town-wide employment lands vacancy rate of 25% is "considered to be too low to account for lands that are marginal (as assessed by the private sector) to develop and/or to provide sufficient variety of choice (size, zoning, location, etc.) to secure new industry. Increasing the vacancy ratio back to the percent level as was the case in 1996 [38.9%] is a more appropriate goal." In a study for the Town of Oakville, Hemson Consulting assumed that 10% of employment lands in Oakville south of Dundas Street would remain vacant over the long term due to "the locational and physical characteristics of the land, the financial situation of the owner, [and] the legal status of the property" (2003e:10). Therefore, a vacancy factor is an input assumption to the model.

3. Jobs located in mixed-use settings. Given the expansive definition of employment lands in this project, this category refers primarily to jobs in public facilities such as schools, as well as jobs in residential buildings such as ground-floor retail and maintenance services. Small-scale street-oriented retail plazas embedded in neighbourhoods may be considered part of this category, but not shopping malls and "big-box" power centres. The number of jobs in mixed-use settings is calculated in proportion to the number of jobs on employment lands.

Step 5: Calculate densities

Population, employment, population-plus-employment, and dwelling unit densities are calculated on the following land bases:

• Gross land area
• Gross land area exclusive of natural heritage features
• Developable land area
• Net residential parcel area
• Net employment parcel area

Summary of model input assumptions

Summary

The inputs to the model were determined through primary and secondary research. This included manipulation of census data and review of plans, academic literature, and consultant reports. Fig. C.5 summarizes the input variables to the model, the units in which they are expressed, and the data on which values were determined. The data on which the input values are based are described in detail later in this appendix.

Fig. C.5: Summary of input variables and data sources

Figs. C.6 summarizes the assumptions that make up the different scenarios. Blank cells indicate that for the scenario in question, Baseline assumptions hold.

Fig. C.6: Summary of scenario assumptions

Undevelopable land

Natural heritage features

The model distinguishes between natural heritage features and the natural heritage system. When rural land is developed, natural heritage features are often assembled into a "system," including corridors for wildlife movement and buffer areas to protect watercourses and wetlands.

Natural heritage systems designated during the development process can account for a substantial proportion of the gross land area to be developed. In the Central Pickering Development Plan, for example, the natural heritage system accounts for 54% of the development planning area (MMAH 2006:32). In its projections for future urban areas, the City of Vaughan's draft OPA 600 (2000: appendix C) assumes that 22.2% of the gross area is "undevelopable."

n As described on pages C-2 to C-5, three hypothetical land bases reflecting different levels of natural heritage protection are defined. All scenarios use the same values for natural heritage features and system land coverage except for the Green scenario, which increases the size of the natural heritage system by 20%.

Developable land

Employment land

As noted, for the purposes of the model, the traditional definition of "employment land" has been expanded to include single-use retail areas. While there may be a reason to distinguish between single-use retail areas and office or industrial parks for policy purposes, they are similar in urban form terms: low, horizontal buildings located on highway-oriented sites with large amounts of surface parking and loading space. Power centres and office and industrial parks also occupy a similar land base. Almost 60% of big-box retailers in the City of Toronto are located on land formerly zoned for industrial use (Jones & Doucet 2000:245). As a result, we can expect most non-home employment to be located on employment lands as defined for the purposes of this model. It can also be argued that single-use retail areas and business and industrial parks produce similar transportation behaviour. Both are automobile-oriented and tend to be located near highways.

Section 2 showed that land for industrial, commercial, and major office uses is distributed unevenly across the metropolitan region. Consequently, there is no "typical" amount of employment land (however defined) at the neighbourhood, district, or even municipal scale. In the three post-1980 study areas containing or near highways, employment land accounted for 11% to 16% of the developable land area. In the other two, employment land accounted for 2%. By contrast, the pre-1960 Riverdale, Oshawa, Oshawa West, and Whitby study areas contained 10% to 22% employment land, with an average of 11%.

• The Baseline scenario assumes that employment land accounts for 10% of developable land area. Given the government's policy preference for a more mixed urban environment in which local area employment-to-population ratios are higher, the Mixed Use scenarios assume that 25% of developable land area is taken up by employment land.

Rights-of-way

The proportion of developable land area accounted for by rights-of-way in the Section 2 study areas ranged from 18% to 35%. There was no discernable correlation between era of development and amount of right-of-way coverage. In the 1980s-90s study areas, between 20% and 29% of developable land was covered by rights-of-way, with an average of 26%. By comparison, in its projections for future urban areas, the City of Vaughan's draft OPA 600 (2000: appendix C) assumes road coverage of 18.5% of the gross area, or 23.8% of the developable area.

• The share of developable land area taken up by rights-of-way in the Baseline scenario is set at 26%. The Consolidated scenario sets a lower amount of road coverage: 20%, a 23% decrease.

Public facilities (parks and schools)

For the purposes of the model, public facilities are defined as schools, parks, and community centres. Libraries, hospitals, and postsecondary educational institutions serve a wider catchment area than the study area, and are therefore removed from the analysis. Land allocation for public facilities was performed using several methods: a pro rata calculation derived from background research for the Central Pickering Development Plan (MMAH 2006); standards contained in the Planning Act; and standards in existing municipal official plans.

As part of the preparation of the Central Pickering Development Plan (MMAH 2006), the land requirements for public facilities to serve a forecast population of 69,000 were estimated. These calculations yielded an overall public facilities dedication of 2.6 hectares per 1,000 persons (planningAlliance, n.d.). This dedication includes parks, schools, places of worship, a library, a cultural centre, health and long-term care facilities, and fire and police stations. Parks and schools account for approximately 90% of the dedication's land area.

A comparison of Section 2 cases revealed that combined park-plus-school land area ranges from 0.98 to 4.81 hectares per 1,000 residents across all cases, with an average of 2.67. (This excludes the Bronte study area, which has an anomalous value.) The post-1980 study areas range from 1.06 to 3.93, with an average of 2.95. (If the Richmond Hill case, which has no schools, is excluded, the average is 3.42.)

The Ontario Planning Act sets maximum standards for parks conveyances that can be required as a condition of development:

42.(1) As a condition of development or redevelopment of land, the council of a local municipality may, by by-law applicable to the whole municipality or to any defined area or areas thereof, require that land in an amount not exceeding, in the case of land proposed for development or redevelopment for commercial or industrial purposes, 2 per cent and in all other cases 5 per cent of the land be conveyed to the municipality for park or other public recreational purposes. ... (3) Subject to subsection (4), as an alternative to requiring the conveyance provided for in subsection (1), in the case of land proposed for development or redevelopment for residential purposes, the by-law may require that land be conveyed to the municipality for park or other public recreational purposes at a rate of one hectare for each 300 dwelling units proposed or at such lesser rate as may be specified in the by-law.

Fig. C.7: Official plan standards for parks

Fig. C.8: School standards, Central Pickering Development Plan

The official plans of Oshawa, Oakville, Brampton, Vaughan, Guelph, Whitby, and Hamilton were also surveyed. Most plans set targets per 1,000 residents. Some also set standards for minimum sizes and population and area served for a hierarchy of parks. (See Fig. C.7.) There is significant variation in these values and the way they are presented. When land for each park type is expressed in terms of land per 1,000 population and summed, the result is a range of values similar to, but in general higher than, that employed in the Central Pickering Development Plan.

Fig. C.8 shows the school standards specified in a background study for the Central Pickering Development Plan (MMAH 2005d).

A survey of reports and official plans reveals that these land area standards per school are consistent with existing practice:

• Vaughan's draft OPA 600 (2000: Part B, s. 4.2.4.2 (v-vi)) sets land areas of 2-3 hectares for elementary schools and 6-7 hectares for secondary schools.
• Whitby's Official Plan (1995: s. 4.7.3.12) states that generally, elementary schools should have a site of 3.0 hectares, though if the school is next to a local or district park, the minimum site size can be reduced to 2.5 hectares.
• A memorandum by Hemson Consulting (2003c) on public facility needs for North Oakville states that for public schools, "assuming that ... schools are located adjacent to an active municipal park, 2.4 ha is required for each elementary school and about 5 ha for a secondary school site." For Catholic schools, "elementary schools require about 3.2 ha or if adjacent to an active park, 2.4 ha."

The Baseline scenario assumes that parks will be allocated to (a) the Central Pickering value of 2.6 hectares per 1,000, or (b) the Planning Act conveyance standards for parks, whichever is greater, plus the following values for schools:

The Consolidated scenario assumes that efficiencies can be achieved through dual-use facilities, either by combining parks and schoolyards or by including parts of the parks system and schoolyards within the NHS. A 1999 report on planners' attitudes towards alternative development standards for public facilities cited a Peel Region task force report that "found that combining reduced road right of way on the residential streets in the 187-acre subdivision, and reducing school site size by one-third, achieved the land dedication required to provide a school site" (Pomeroy 1999:7). The report also noted that

...combining community facilities such as schools and parks can provide up to a 15 per cent reduction over the cost of segregated facilities. Similarly, utilizing park and open space dedications as part of a storm water management system can combine dedications and increase efficiency of land use. This has been achieved in ... Markham and Ajax. (7)

• The Consolidated scenario reduces the allocation standards for public facilities by 20% relative to the Baseline scenario.

Fig. C.9: Summary of recent and forecast housing production by unit type

a. Source: Will Dunning Inc. (2004) 5.

b. Source: Hemson (2005) appendix E.

The "905" area refers to the Regional Municipalities of Halton, Peel, York, and Durham.

Population and housing

Housing type mix

The mix of housing unit types will depend on housing affordability, interest rates, local and provincial policies, and demographic change. As housing becomes more expensive, demand will shift away from detached housing and toward less expensive housing types, such as attached dwellings and apartments (Will Dunning Inc. 2006). Fig. C.9 summarizes recent and potential housing growth by unit type.

• The Baseline scenario assumes the continuation of the 1999-2003 housing type mix for the "905" area. The Forecast Mix scenario assumes the Hemson "Compact" forecast housing type mix. The Market Shift scenario assumes the Hemson "More Compact" forecast housing type mix. In Forecast Mix and Market Shift scenarios, it is assumed that 20% of apartment units are in stacked townhouse form.

Average household size by housing type

According to the 2001 Census, the average household size in the "905" area was 3.11 persons. This is higher than that in established urban centres such as the Cities of Toronto and Hamilton, which are 2.63 and 2.61, respectively. Larger household sizes in newer areas are the product of both demographic and spatial factors. Fig. C.10 shows the forecasts for average household size by unit type assumed in the Central Pickering Development Plan (Will Dunning Inc. 2004; MMAH 2004:34). In its Visualizing Density study, the Region of Waterloo assumes slightly lower average household sizes: 2.94 for single and semi-detached units, 2.69 for townhouses, and 1.85 for multiple dwelling units (2007:11).

• All scenarios adopt the Central Pickering values.

Residential parcel area by unit type and units per lot

Fig. C.11 shows the assumptions for average residential parcel area by housing type used in the projections for the Central Pickering Development Plan (Will Dunning Inc. 2004). Comparisons to studies of built form and density at the parcel scale show these values to be consistent with measurements in the GTA and elsewhere (see Fig. C.12.). See also Design Center (n.d.) and Campoli & MacLean (2007).

Fig. C.10: Forecast household size by housing type

Fig. C.11: Average residential lot area by housing type

Fig. C.12: Comparison of net residential densities by housing type

All scenarios assume the Central Pickering density values for each unit type.

The Central Pickering apartment form is consistent with design samples in the Regional Municipality of Waterloo (2007) and BLGDG (2005) reports. The Capers Block, a Vancouver model used in the Waterloo study (2007:97), has a lot area of 0.52 hectares, is five storeys high, and contains 78 units, for a net parcel density of 150 units per hectare. The BLGDG study models are situated in denser urban contexts, resulting in smaller lot areas. The Market Square, a Toronto model, is eight storeys and contains 306 units on a 1.17-hectare lot, for a net parcel density of 262 units per hectare. The Central Pickering apartment model represents an intermediate height and density. It is assumed that higher-rise apartments are unlikely to locate in greenfield neighbourhood areas; such development is more likely to be channelled to planned nodes, especially the "urban growth centres" specified in the Growth Plan.

Employment

Employment location

Consultants typically divide employment into four categories: jobs located in the home, jobs located in freestanding office buildings, population-related employment (retail, education, and services embedded in neighbourhood areas), and jobs located on "traditional" employment lands (industrial, commercial, warehousing, and offices in business and industrial parks).

Employment on segregated employment lands

A survey of recent planning reports in the Toronto region indicates that the more recently a municipality has been developed, the higher the proportion of its workforce employed in business and industrial parks. In the "905" area as a whole, 55% of jobs are on "traditional" employment lands (business and industrial parks), while the figures for the Cities of Toronto and Hamilton are 31% and 43%, respectively. The City of Vaughan is the highest, at 69%. When jobs in major office and in business and industrial parks are combined, the total is approximately two-thirds in the "905" area.

The remaining third -- populated-related jobs -- are largely in the retail, education, accommodation, health, and arts and entertainment sectors, and in the home. In newly developed areas, retail jobs tend to be located on segregated, single-use parcels disconnected from the residential urban fabric: shopping malls, power centres, and in business parks. If retail jobs are added to the business and industrial parks and major office categories, then over three-quarters of jobs are located on parcels segregated from the residential neighbourhood fabric.²¹ (See Fig. C.13.)

Fig. C.13: Estimates of present job location (excluding jobs in the home)

The right-hand column is a proxy for employment in segregated, single-use employment zones. Columns B and D are overlapping categories, and so columns A, B, C, and D do not total to 100%.

a. All but Oakville, Inner Ring outside Toronto, and GTA from Hemson (2003d:10). Oakville value from Hemson (2003e:9). GTA value from Hemson (2003b:33-34; Lorius 2004).

b. Census 2001 Place of Work data employment by NAICS code.

c. Aggregate retail trade sector employment for Brampton, Markham, Milton, Mississauga, Oakville, Oshawa, Pickering, Richmond Hill, Vaughan, and Whitby.

Fig. C.14: Forecast location of future employment growth, 2001-2031

Source: Hemson (2005). Calculated from Appendix F, Compact Scenario.

In the future, Hemson Consulting forecasts that, for the "905" area as a whole, 20% of the additional jobs will be located in free-standing office buildings, 50% in business and industrial parks, and 30% elsewhere. The percentages vary among upper-tier municipalities. (Hemson 2005; see Fig. C.14.)

• In the Baseline scenario, 82% of all jobs are on employment lands: 20% in major office, 50% in business and industrial parks, and 12% in single-use retail zones such as shopping malls and big-box power centres.

Fig. C.15: Overall job mix

Overall job mix

Within employment lands, jobs are assigned to three categories of land: business and industrial parks, major office, and single-use retail areas. Fig C.15 shows total employment by location for the Baseline scenario, as well as the "905" area forecast from Fig. C.14.

Up to this point, the analysis has neglected home-based employment. Section 2 showed that since employment land tends to be congregated in large-scale zones separated from residential areas, there tends to be little employment land -- and therefore few jobs -- in recently constructed neighbourhoods. For this reason, calculating the number of jobs in mixed-use settings as proportion of total jobs will likely produce an underestimate. On average in the Toronto, Hamilton, and Oshawa CMAs, about 6% of all members of the employed labour force work out of their homes. On this basis the number of jobs in mixed-use settings is topped up by adding 6% of the employed labour force, assuming a labour force participation rate of 0.60, to the number of jobs in mixed-use settings.

Some of the jobs in business and industrial parks are compatible with mixed-use settings. Fig. C.16 shows the composition of employment in business and industrial parks in Vaughan, Mississauga, and Markham. In Vaughan and Mississauga, about 20% of jobs in business and industrial parks are in the business and personal services sectors. In Markham, it is almost half. In Vaughan and Markham, a further 6% of jobs in designated business and industrial parks are in retail trade.

A note on job mix, geographic scale, and population-related employment

A few comments on the sketch model's treatment of employment are in order. First, the overall job mix is derived from forecasts at the municipal scale. On this basis, it is assumed that, to some degree, municipal proportions will be replicated at smaller geographic scales -- in this case, the 2km-by-2km scale. In today's urban development patterns, this does not occur, but if more "complete communities" are built, a broader range of employment would be found at the district scale.

Second, the model calculates employment in mixed-use settings in proportion to the number of jobs on employment land, which is determined earlier in the process. This is a convenience. In a land-optimizing model, "population-serving employment" is typically determined in proportion to the resident population. For example, the Central Pickering background report on employment land notes (without citation) that "the accepted standard is 1 job for every 5 persons" (MMAH 2005a). In an activity-optimizing model at the submunicipal scale, it cannot be assumed that such a ratio will hold.

Fig. C.16: Jobs in business and industrial parks by sector in Vaughan, Mississauga, and Markham

a. Personal Services combines NAICS categories Information; Culture and Recreation; Accommodation and Food Services; and Other Services. Business Services combines Professional, Scientific, and Technical Services; Management of Companies and Enterprises; and Administrative and Support, Waste Management & Remediation Services.

b. Source: Regional Municipality of York (2003).

c. Source: City of Mississauga (2005a). Employment districts included are: Mavis-Erindale, Dixie, North East, Southdown, Airport Corporate, Sheridan Park, Gateway, Meadowvale Business Park, and Western Business Park. Pearson Airport and Downtown are excluded. These areas represent 95% of designated employment land in the City of Mississauga.

The methodological problem is this: unlike employment land, where total jobs can be derived from land area using density parameters, a "supply-side" approach cannot by definition be used in mixed-use settings. An alternative approach would be to detach employment land from jobs in mixed-use settings, and somehow calculate the latter in proportion to resident population. This would require substantial additional research into the nature of such employment; such research is beyond the scope of this study. As an experiment, the number of jobs in the NAICS "education" category was quantified for the 15 districts analyzed in Section 2 that contained schools, revealing that between 4% and 16% of total employment was education-related, with an average of 9% -- half the value for jobs in mixed-use settings in the Baseline scenario. In the end, the population-to-employment ratios produced by the model (8 to 8.9 in all but the Mixed-Use and Jobs-Housing Balance scenarios) are within the range found in four of the five post-1980 cases analyzed in Section 2 (5.27 to 12.22).

• The Mixed-Use scenario assumes that, relative to the Baseline scenario, half of retail, business, and personal services jobs in business and industrial parks will shift to mixed-use settings. Assuming that approximately one-third of jobs on employment lands falls into these categories, half would amount to a 15 percentage point shift. It is also assumed that an additional 10% of jobs in business and industrial parks -- 5 percentage points -- will shift to the major office category. The Mixed-Use and Baseline scenarios are summarized in Fig. C.17.

Fig. C.17: Jobs location in the Baseline and Mixed-Use scenarios

Employment density

Although the Ontario government focuses on measuring density in terms of jobs, land use planning for employment is typically concerned with built form characteristics: Gross Floor Area (GFA), Floor Area Ratio (FAR), and -- especially in office settings -- interior floor space per worker. In addition, developers distinguish between gross and net floor space, or the total floor area of a building versus the area net of walls, elevators, corridors, and other common or utility areas.

A study by Ove Arup (2001) on employment density measurement for the British government found that floor space per worker varied considerably depending on the location of the building, its age, the nature of the job, the sector of the employer, tenure, and even position in the business cycle. The study did not approach the difficult issue of space external to the building itself, for example for parking, internal roadways, or mandated greenspace.

The U.K. government guidance for local authorities on best practices for use in review of employment lands proposes a multi-stage process to convert job type to gross parcel area per job (Office of the Deputy Prime Minister 2004: Annex D):

number of jobs

• net interior floor space per job
• net interior floor space to gross floor space (net-to-gross ratio)
• gross floor space to parcel area (plot area ratio)

= total land requirement

Given the lack of Toronto-area data and the variations in each variable at each step, this approach was rejected. Instead, average job densities for each employment land type were derived from available information.

Density of office and industrial employment

The net density of jobs on designated office, commercial, and industrial employment lands has been documented in consultant reports for several municipalities:

• In 1996, Oakville's job density on employment lands was found to be 17 jobs per acre, or 42 jobs per hectare (Oakville Economic Development Alliance 2000:10).
• A 2002 report for Halton found that: "Currently, Halton's employment density is about 45 employees per net hectare. This figure is similar to Vaughan, somewhat lower than Mississauga, but higher than Brampton which is below 35 employees per net hectare" (Hemson 2002:19).
• A 2002 report for Burlington found an average of 35 manufacturing and construction jobs per hectare; transportation, storage, communication, utilities, education, health, accommodation and food averaged 40 per hectare; trade-related jobs averaged 50 per hectare; and finance and business services averaged 100 per hectare (Metropolitan Knowledge International et al. 2002:10-11).
• Background studies for the Central Pickering Development Plan found "employment densities in mixed industrial/office business parks in the GTA average around 40 jobs per net hectare [and] higher density office centres average roughly 100 jobs per net hectare in these communities" (MMAH 2005a).

Profiles prepared by the City of Mississauga Economic Development Office for nine employment districts indicate that the eight areas in which manufacturing, wholesale trade, and transport constitute the majority of employment activity have developed area densities of between 10 and 55 jobs per hectare, with an overall value of 43 jobs per hectare. If Southdown (10 jobs per hectare) is excluded, the density range is 37 to 55 jobs per hectare. The land base for this calculation is composed of non-vacant parcels. It is unclear to what extent this includes internal roads, public open space, and other forms of ancillary land use. If so, the net parcel densities would be slightly higher.

When calculated on a gross basis -- i.e., including undeveloped parcels -- the density ranges from 8 to 43 jobs per hectare, with an overall value of 35. This difference between the gross and developed area densities is accounted for by the fact that the employment lands are, overall, about 80% occupied. A ninth district, Airport Corporate Centre, is dominated by office-format employment. This area has a developed area density of 137 jobs per hectare at a gross density of 105. Together these nine employment districts account for 95% of all employment land in the City of Mississauga, and are therefore representative of the City as a whole (City of Mississauga 2005a,b).

Generally speaking, the higher the proportion of manufacturing, wholesale trade, and transportation and warehousing, the lower the density. This is as expected, given the land consumptiveness of these activities. (See Fig. C.18.)

For comparison, Fig C.19 shows net densities for industrial and office employment taken from Nelson (2004).

Fig. C.18: Jobs, land area, and density of Mississauga employment districts (2005)

Source: City of Mississauga, "Business..." (2005). Mfg = Manufacturing; Wh = Wholesale trade;

Tr = Transportation and Warehousing; Pr = Professional, scientific, and technical services.

Fig. C.19: Net densities of industrial and office employment

Source: Nelson (2004:47).

These values do not account for vacant land and therefore represent only developed land area. Nelson also determines average densities for industrial and office employment by multiplying the job density for each type by the projected share of the labour force accounted for by each type, resulting in densities of 55 and 116 employees per net hectare, respectively. These values are slightly higher than, but comparable to, the values in the GTA consultant reports.

• In the Baseline scenario, office and industrial business parks with some office component are assumed to achieve a density of 40 jobs per hectare. Higher-density office centres are assumed to have a density of 100 jobs per hectare. The Mixed-Use scenario assumes that, relative to the Baseline scenario, the density of office and industrial business parks will increase by 25%, to 50 jobs per hectare, and the density of major office will increase from 100 to 125 jobs per hectare. For all scenarios, employment lands are assumed to be 20% vacant.

Density of retail employment

The density of single-use retail areas in the Toronto region has not been studied. Due to the vast differences in workforce required to support, for example, a mall filled with small boutiques versus a "big-box" superstore, as well as the different parking needs for different types of retail facilities, there is no "typical" density. The Ove Arup (2001) study determined densities for "town centre" (inner-city retail strips), food superstores, and warehouse-style big-box retailers. All were expressed in terms of employees per internal floor area rather than in terms of gross site density. For example, food superstores were assigned an average density of 19 m² of net internal floor area per worker, while warehouse-style big-box stores were assigned an average density of 90 m² of gross internal floor area per worker.

Nelson (2004) found that, after accounting for vacancy rates, in neighbourhood shopping centres serving a local population of 3,000 to 40,000 people, each employee occupies 632 ft² (59 m²) of gross floor space. He assumes an FAR of .23, resulting in a density of 39 jobs per hectare on the gross site area. Assuming, as he does, that a neighbourhood shopping centre occupies 3 to 10 acres (1.2 to 4 hectares), a typical shopping centre facility would contain between 50 and 150 workers. Nelson also derives densities for other, larger shopping centre types. (See Fig. C.20.)

Fig. C.20: Density of shopping centres

Source: Nelson (2004:43-47).

Fig. C.21: Characteristics of three GTA shopping centres

a. Employment numbers are taken from the Census, aggregating the Retail, Administration and Support, and Accommodation and Food Services NAICS categories.

To obtain baseline data for the Toronto region, the property managers of four shopping centres in the Greater Toronto Area were contacted. As much as possible, sites were selected that aligned with and were the sole employer in a census dissemination area. The managers of each site were asked for the gross land area of the site, the estimated number of employees, the area of the building footprint, land area for parking, number of parking spaces, leasable retail floor area, and the year the facility had originally been developed. The number of employees by NAICS code was taken from the Census and compared to the jobs total applied by the site manager. (See Fig. C.21.)

The results were inconclusive. The property managers of only three of the four sites were willing to share information: two outer suburban malls constructed in 1986 and 1991, and a recently renovated 1960s-era mall. The two suburban malls of comparable site area and worker population were found to have gross employment densities of approximately 80 jobs per hectare. It seems that all three sites have higher floor space per job and gross site density than those suggested by Nelson. The underlying variables differ significantly, however, making it difficult to generalize from these cases with confidence.

Big-box superstores are not included in this analysis. Due to the fragmented management of power centres, obtaining land use and employment information was not attempted. Given the similarity in built form, these densities may be comparable to warehousing facilities. This assumption is partially corroborated by Ove Arup (2001), which found that gross internal floor space per worker of big-box superstores (90 m²) and large-scale warehousing and distribution facilities (80 m²) are similar. These values are higher than those for shopping centres from Nelson and the three-site study, suggesting that jobs in big-box superstores occupy more floor area.

• Given limited information and resources, gross employment density for single-use retail areas must be inferred. A value of 50 jobs per hectare was used for the Baseline scenario. In the Mixed-Use scenario, retail density is expected to increase by 25% to 62.5 jobs per hectare.

Jobs-housing ratio

A person who lives near employment opportunities at least has the option of walking or cycling to work. If residents choose to work locally, local-area residential-employment balance would result in "self-containment" and a reduction in the number and length of commuting trips.

Census data show that overall, the municipalities of Oakville, Markham, Richmond Hill, and Vaughan each have labour force participation rates of 0.60, meaning that for every 100 residents, 60 are members of the employed labour force. Mississauga's is 0.61 and Milton's is 0.65. If the participation rate is 0.60 in all scenarios, jobs-housing balance would exist if there were 1.66 residents for every job.

• In the Jobs-Housing Balance scenario, it is assumed that the number of jobs within the study area is equal to the number of residents who are members of the employed labour force over the age of 15, or 1.66 residents per job.

Summary of model outputs

As discussed in Appendix C.2, several formulas were used to calculate land allocations for public facilities. In each case, the combination of formulas used to calculate the reported values was:

• The sum of the Planning Act s. 42(1 & 3) parkland dedication of one hectare per 300 dwellings and 2% of employment land; plus
• The official plan standards for schools, which allocated three classes of schools, each with different land areas per institution, in proportion to the number of dwellings.

Fig. C.22 summarizes the outcome for each natural heritage protection case and scenario: the amount of public facilities land per thousand people, the number of schools by class of institution, and the absolute number of people, jobs, and dwellings. Figs. C.23-C.25 show densities on all land bases for each scenario and natural heritage protection case.

Fig. C.22: Comparison of population, dwellings, employment, and public facilities

Fig. C.23: Comparison of scenario densities - Low

Fig. C.24: Comparison of scenario densities - Medium

Fig. C.25: Comparison of scenario densities - High

Notes

21. For simplicity's sake, this assumes that no retail jobs are located on "traditional" employment lands. York Region staff estimate that 25% of retail jobs are located in business parks (personal correspondence; see also Fig. C.16). If retail jobs make up approximately 12% of total employment, then the proportion located in business parks could equal 3% of total employment.