Scotland-wide Older and Disabled Persons Concessionary Bus Scheme - Further Reimbursement Research
Summary
Context
Under the Scotland-wide Older and Disabled Persons Concessionary Bus Scheme, qualified Scottish residents may travel anywhere in Scotland at any time on local buses at no charge. Transport Scotland is responsible for compensating bus operators for carrying these passengers. It has the objective that bus operators should be financially "no better off and no worse off" because of their participation in the scheme, requiring it to identify a "counterfactual" situation against which to measure:
- the revenue that would have been earned by the operator if passholders were not able to travel for free ("revenue forgone"); and
- any additional costs that the operator might have incurred as a consequence of providing free travel, in particular through carrying passengers who would not otherwise have travelled.
In 2012-13, nearly 150 million free concessionary passenger journeys are expected to be made, at a cost to the Scottish Government (and therefore taxpayers) of £187 million in reimbursement payments under current arrangements.
Reimbursement payments are calculated by applying a National Rate to the adult cash single fare that would have been paid for individual concessionary journeys. The National Rate was initially fixed at 73.6% in 2006, but was reduced to 67% with effect from April 2010. However, the budget for the National Concessionary Travel Scheme is capped to £187m in both the 2012-13 and 2013-14 financial years.
Although reimbursement rates have been informed by various research studies, Transport Scotland has not had a systematic method for updating the Reimbursement Rate to reflect changes in commercial fares and relevant cost measures. The objective of the current study was to provide Transport Scotland with an appropriate mechanism for doing so.
Given the hypothetical nature of the counter-factual, and the large sums of money involved, determining appropriate reimbursement levels is inevitably a matter of considerable debate between the authorities responsible for paying it and the operators who receive it.
This study has been carried out with the active co-operation of the Confederation of Passenger Transport - Scotland (CPT) on behalf of Scottish bus operators, with CPT's advisers Steer Davies Gleave (SDG). For brevity throughout this Report we use 'CPT' as shorthand for the combination of CPT and SDG. We are most grateful for their participation.
Although the data and analysis methods that we and CPT have used are largely identical, there remains room for alternative interpretations and assumptions. Although our focus is on the results that we believe are most robust, where relevant we also identify our understanding of the alternative conclusions that CPT have reached, and quote compromise values for key parameters that recognise the uncertainties associated with the analysis.
Average fares
One of the key factors in determining "no better off and no worse off" reimbursement is the average fare that would be paid in the absence of the concession. Reimbursement is calculated by reference to the "shadow fare", that is, the adult cash single fare that would have been charged for individual passholder concessionary journeys. However, if there was no concession, passholders are likely to use some combination of the different tickets promoted by bus operators, including day tickets and weekly tickets. Typically these will enable journeys to be made more cheaply than using cash single fares, and consequently a discount factor needs to be applied to convert the "shadow fare" to a better approximation to the average fare per journey that would have been paid in the counter-factual.
CPT has provided access to data from bus operators on commercial ticket sales and revenues of non-concessionary passengers, and estimates of the passenger journeys made using these tickets. The average fare paid by non-concessionary passengers can therefore be calculated, and by making assumptions about how their travel patterns would differ from those of concessionary passengers, the average fare that would be paid by concessionary passholders can be estimated.
It has been possible to confirm the general scale of the CPT assumptions about the journeys made per ticket using Transport Scotland data, and we are content that the resulting estimate of the discount factor is reasonably robust. Our analysis suggests that the discount factor should be 19.25% (i.e. that the average fare that would be paid by passholders in the counterfactual is about 81% of the average shadow fare), which contrasts to the value incorporated into the current National Rate of 7.3%[1]. The new analysis therefore suggests a significant increase in the amount of discount applied, leading to reduced reimbursement payments, other things being equal. CPT's conclusion from analysis of the same data is that the discount factor should be 18%. A compromise value for the discount factor is 18.6%.
The methodology leading to these values relies upon data that has been collected voluntarily by CPT from its members. We recommend that future updating of the discount factor estimate is driven by data collected from bus operators through a more formal process which gives Transport Scotland more assurance about the quality and completeness of the information provided. We have also developed a methodology that could be used to avoid dependence upon arbitrary assumptions in this updating process.
Reimbursement Factors and Elasticities
A central feature of the calculation of "no better off and no worse off" reimbursement is the concept of "generation", which is that a proportion of concessionary journeys would not have been made by passholders in the absence of the zero fare. The Reimbursement Factor is used to measure the scale of generated journeys, and to infer the number of concessionary journeys that would have been made in the counter-factual. If the fare that would have been paid in the absence of the scheme increases in real terms, then counter-factual journeys could be expected to reduce, and, correspondingly, the proportion of generated concessionary journeys to increase.
The sensitivity of demand to fares is usually measured in terms of elasticities. The Reimbursement Factor in the current National Reimbursement Rate was established in 2006 from predecessor schemes, but the evidence base that supports the current implicit elasticities is weak. Elasticities are most easily identified when a large change in fare allows the resulting change in journeys to be measured. To provide a robust basis for reimbursement arrangements going forward, we have estimated new concessionary travel elasticity values for Scotland by using data from before and after October 2002, when statutory free concessionary travel was introduced in Scotland.
Our analysis has drawn on a wide range of data to ensure that comparisons of journeys are on a like-for-like basis, but there are challenges in assembling the necessary information from ten years ago. The most complete and precise information is available for the Strathclyde area, but we have had to make a variety of assumptions to enable Scotland-wide estimates to be made. There is room for alternative interpretations of the available data, and CPT has developed their own elasticity estimates.
Our preferred values (which give a point elasticity at 2001-2 fares and prices[2] of -0.32) are based on long-run analysis of all-Scotland data, and are larger in absolute terms than the values implied by the current Reimbursement Factor as derived from fares in 2006-7 (-0.29). CPT's estimates give a point elasticity of -0.27, slightly larger in absolute terms than those implied by the current Reimbursement Factor and current fare levels. A compromise value gives a point elasticity of -0.29 at 2001-2 fares and prices.
The 2012-13 Reimbursement Factor implied by our preferred elasticity value (for illustrative purposes, assuming the "compromise" discount factor) is 58.2%, which is somewhat lower than the 61.5% in the current Net Reimbursement Rate. The equivalent Reimbursement Factor implied by the CPT estimate is 62.4%, which is slightly higher than the current Reimbursement Factor. The compromise elasticity implies a Reimbursement Factor of 60.3%.
In calculating Reimbursement Factors on an on-going basis, it is necessary to remove general price inflation from changes in fare levels. Usually the Retail Price Index (RPI) or Consumer Price Index (CPI) is used, but we have also looked at alternatives which reflect the significance of petrol prices in determining passholder mode choice. We recommend a combined index which gives additional weight to the Petrol and Oil component within the RPI. We have used survey data which quantified the extent to which the concession encourages switch of mode from car to bus to derive a 22% weighting for the Petrol and Oil component.
Additional Costs
A previous study by ITS[3], reporting in 2010, undertook detailed analysis of the additional costs likely to be incurred by operators in providing the concession. We have largely taken the conclusions from the previous work as a sound basis for future reimbursement, reflected in the ITS recommendation that reimbursement for additional costs is calculated on the basis of a cost rate per generated concessionary passenger.
However, we have examined a number of issues raised by CPT regarding the way in which differences in operating costs levels between Scottish bus operators and the rest of Great Britain were handled by the ITS Study. A detailed reconstruction of the ITS analysis is not possible within the available time and resources, but we have concluded that there is justification for increasing the base cost rate recommended by ITS from £0.344 per generated passenger to £0.428 in 2009-10 prices. We understand that CPT believe that there is a case for further increasing this rate to £0.459 in 2009-10 prices, but we do not think there are grounds for doing so. A compromise value is £0.443 per generated passenger at 2009-10 prices.
In applying the additional cost rate, it is necessary to update values to reflect changes in prices and costs actually experienced by Scottish bus operators. CPT publishes a well-established index of bus and coach industry operating costs which includes separate values for Scottish operators. We recommend that this index is used for updating the base cost from 2009-10 prices to the period for which future reimbursement calculations are being made.
Implications of reimbursement parameters
The work summarised above has led to the identification of a set of reimbursement parameters (our "preferred" values) which in our view represent the most robust basis for calculating future reimbursement payments. Overall, our conclusion is that the evidence suggests that the reimbursement arrangements of the current National scheme are over-generous to bus operators and should be reduced.
In some cases, different values have been identified by CPT, reflecting different judgements about some of the assumptions and alternative interpretations of evidence, and suggesting that somewhat higher levels of reimbursement are justified relative to our preferred values. As noted above, we have also identified a set of "compromise" parameters that are simply a straight average of our preferred estimates and CPT's estimates where these differ. Both our preferred and compromise set of reimbursement parameters imply reduced levels of reimbursement payment compared to the current National Scheme.
The table below summarises the three main alternative sets of reimbursement parameters, and illustrates their implications for reimbursement payments, using predicted values of 2012-13 concessionary journeys and the average shadow fare, based on the first 9 periods of 2012-13.
| Reimbursement Parameters | MVA/Minnerva Preferred Values | CPT | Compromise values |
|---|---|---|---|
| Reimbursement Year | 2012-13 | 2012-13 | 2012-13 |
| Concessionary journeys, (m) Reimbursement Year | 146.568 | 146.568 | 146.568 |
| Average concessionary (shadow) adult cash single fare, Reimbursement Year prices | £2.041 | £2.041 | £2.041 |
| Discount Factor relative to Shadow Fare | 19.25% | 17.96% | 18.60% |
| Average fare that would have been paid in the absence of the scheme (2001-2 prices) | £1.10 | £1.12 | £1.11 |
| Lambda | 0.703 | 0.673 | 0.688 |
| Beta (Deindexed, at 2001-2 prices) | -0.503 | -0.439 | -0.471 |
| Illustrative point elasticity at 2002 fare of £0.859 | -0.318 | -0.267 | -0.292 |
| Reimbursement Factor | 58.4% | 62.3% | 60.3% |
| Revenue forgone | |||
| Reimbursement for revenue forgone (m) | £141.05 | £152.87 | £146.82 |
| Additional costs | |||
| Base marginal operating and capacity cost per generated passenger, 2009-10 prices | £0.428 | £0.459 | £0.443 |
| CPT Cost index relative to 2009-10 | 117.0 | 117.0 | 117.0 |
| Cost per generated passenger in the reimbursement year | £0.500 | £0.537 | £0.518 |
| Generated journeys (m) | 60.985 | 55.274 | 58.194 |
| Reimbursement for additional costs (m) | £30.49 | £29.67 | £30.17 |
| Total reimbursement calculated (uncapped) (m) | £171.5 | £182.5 | £177.0 |
| Net Reimbursement Rate (as % of journeys * shadow fare) | 57.3% | 61.0% | 59.2% |
| 2012-13 Predicted Reimbursement Payment | £187.0 | £187.0 | £187.0 |
| Reimbursement parameters as defined in current "scheme" | |||
| "Generation Factor" G | 58.4% | 62.3% | 60.3% |
| "Discount Factor" D | 11.2% | 11.2% | 11.2% |
| "Additional Cost rate" C | 10.2% | 9.9% | 10.1% |
The table shows how the different parameter choices lead to different estimates of reimbursement payments, which vary from £171.5m to £182.5m in 2012-13. The Net Reimbursement Rate similarly varies from 57.3% to 61.0%; these values contrast with the current Rate of 67%, although the impact of the £187 million cap on the total payment reduces the effective rate to 62.8%.
Reimbursement Calculations for 2013-14 and 2014-15
Whatever reimbursement parameters are decided upon, the framework we have used to calculate reimbursement provides a valuable on-going tool which Transport Scotland can use to update reimbursement calculations as and when required. For example, with forecasts of future trends in cash single fare scales, it is possible to calculate the Net Reimbursement Rate that might be applied to reimbursement payments in 2013-14 and 2014-15. With forecasts of trends in concessionary journeys, total reimbursement payments can also be calculated.
The table below summarises the results of these calculations for the different reimbursement parameters, using the predicted 2012-13 outturn figures and assumptions that:
- there will be no change in concessionary numbers to 2013-14 and 2014-5;
- the average adult cash single fare will increase by 5% per year in current prices; and
- prices generally, including petrol prices, and bus operating costs, will increase by 2.5% per year.
| Illustrative forecasts of reimbursement payments and Net Reimbursement Rates | MVA/Minnerva Preferred Values | CPT | Compromise values |
|---|---|---|---|
| 2012-13 | |||
| Concessionary journeys (m) | 146.568 | ||
| Average adult cash single fare (current prices) | £2.041 | ||
| Total reimbursement calculated (uncapped) (m) | £171.5 | £182.5 | £177.0 |
| Net Reimbursement Rate (as % of journeys * shadow fare) | 57.3% | 61.0% | 59.2% |
| 2013-14 | |||
| Concessionary journeys (m) | 146.568 | ||
| Average adult cash single fare (current prices) | £2.143 | ||
| Total reimbursement calculated (uncapped) (m) | £178.4 | £190.1 | £184.2 |
| Net Reimbursement Rate (as % of journeys * shadow fare) | 56.8% | 60.5% | 58.6% |
| 2014-15 | |||
| Concessionary journeys (m) | 146.568 | ||
| Average adult cash single fare (current prices) | £2.250 | ||
| Total reimbursement calculated (uncapped) (m) | £185.5 | £197.9 | £191.6 |
| Net Reimbursement Rate (as % of journeys * shadow fare) | 56.2% | 60.0% | 58.1% |
It can be seen that with these assumptions, the Net Reimbursement Rate calculated from our Preferred reimbursement parameters falls from 57.3% to 56.2%, whereas with the compromise values, the Rate falls to 58.1%.
In determining future levels of the Net Reimbursement Rate, Transport Scotland should review trends in fares and concessionary journey numbers to ensure that inputs into the calculations are based on reasonable assumptions.
Longer Term Issues
We recommend Transport Scotland considers the option of moving away from the use of a single national reimbursement rate to determine the reimbursement payments for individual operators.
In our view, operator payments calculated from a single national rate which is the same for all operators gives rise to a number of issues. At the individual operator level, it may not leave each individual bus operator no better off, and no worse off (i.e. some bus operators will be better off, and others will be worse off), even if it satisfies this objective at an All-Scotland, operator-wide level.
Since reimbursement is largely driven by the cash single fare, a national rate provides a strong commercial incentive for individual bus operators to increase their cash single fares at a greater rate than they would do otherwise. This incentive is greatly increased by the existence of the cap on payments, which gives a direct inducement for operators to increase fares sooner rather than later to maximise their own share of the available pot in any given financial year.
In our view the fact that the cash single fare has risen as rapidly as it has between 2006-7 and 2012-13 is almost certainly a consequence of this weakness.
We therefore recommend that Transport Scotland reviews options for change from the standard national rate, so as to better reflect "no better, no worse off" objectives at the individual operator level, and reduce the incentive for operators to introduce increases to the fares used to calculate reimbursement payments. There are a variety of options available, which reflect different balances between theoretical and practical issues. However, adoption of a different methodology could risk significant unintended consequences for the bus industry, and an important part of an assessment of options should be the evaluation of these risks, alongside the potential benefits.