9.3.5.2 Agglomeration Benefits - APARC Method

Using a combination of the Transport Model for Scotland (TMfS) and its land-use component - Transport Economic Land-use Model of Scotland (TELMoS), it is possible to implement the methodology discussed above in Scotland. The level of disaggregation is to each TMfS zone. This is fairly time consuming and requires, at the very least, the use of TMfS.

An alternative, simpler approach is outlined here - the Agglomeration Productivity Aggregate Response Calculator. It should be noted that this approach does not significantly differ theoretically from that discussed above, but is simply an alternative and less resource intensive application. In practice either approach may be used.

The basic principle of the approach is to define a single zone for the purposes of the calculation.

Applying this assumption means that equations 2 and 3 become:

Where superscripts S and M indicate the do-something and do-minimum cases respectively. Note that the α parameter is not included above as it is assumed that it is equal to 1. The key term is g. This represents, under the single zone assumption, a measure of the generalised cost of travelling in the defined zone. It is helpful to rewrite equation 4 as:

and thus, the terms and represent the ratio of the generalised cost of travel in the base year versus the do-something and do-minimum respectively. This is useful as it means that the precise form of generalised cost is not important - what is important is the proportional change of both the do-minimum and do-something over the baseline. This method has the advantage in that the modelling resource requirements are kept to a minimum - any transport model should, in theory, be able to generate a change in total generalised cost.

It is worth noting that g_ij^X as originally defined is a distance based measure - two zones are clearly a physical distance apart as the crow flies although new infrastructure may mean the distance that needs to be traveled may change. The generalised cost measure in the aggregate case is not distance based in the same way.

The first data requirement is the elasticity parameter, e. The current best estimate of this value is 0.041.

Values are also required for h, the level of GDP per worker in the zone. In order to avoid double counting, this value is kept constant between the do-minimum and do-something scenarios.

It is also suggested that total employment is assumed to be constant between the do-minimum and do-something.

Standard assumptions should be applied in terms of GDP and employment growth from the base to the forecast year or years and onward. For consistency with current appraisal techniques, all values are expressed in 2002 prices and discounted over 60 years from scheme opening using standard Green Book values.

These parameters are fixed and suggested values are given in Table 9.1 below for base years between 2002 and 2006. Note that when TMfS is used to calculate agglomeration benefits, it calculates the generalised cost of travel for Scotland as a whole, and therefore national employment and GVA per worker figures should be used. These are also given in Table 9.1. Practitioners should note that the 2% GVA growth figure will be adjusted over time in line with the Government Economic Strategy.

The method of arriving at a measure of total generalised cost (GC) for each case is as follows:

1. For each mode/time of day a measure of GC per km traveled is calculated.

2. The total measure of GC is calculated, for each case, by multiplying the GC/km figure for each mode and time of day by the appropriate level of vehicle km in the Base.

It should be noted that there may be variations in the amount of information provided into the above calculation. If this is the case then the basic principle outlined should be followed.

This is consistent with the full methodology and captures the fact that agglomeration is dependant on the impact of changes in the transport system compared with the base. It is worth noting that congestion effects, etc., are captured in the case specific GC figures.

Equation 4 should be used to calculate the extent of agglomeration benefits in each of the forecast years. Values should be discounted back to 2002 for consistency with standard appraisal. If a single forecast year is available, then the agglomeration impact should be held constant in subsequent years. If two forecast years are available, linear interpolation should be used to calculate interim values and the agglomeration benefits should be held constant after the second forecast year.

A spreadsheet is available in Section 17.5 that can perform the calculation.