ARTICLE IN PRESS

Transport Policy 17 (2010) 20–26

Contents lists available at ScienceDirect

Transport Policy

0967-07

doi:10.1

n Corr

E-m

journal homepage: www.elsevier.com/locate/tranpol

Impacts of free concessionary travel: Case study of an English rural region

Stuart Baker a, Peter White b,n

a Hampshire County Council, UKb University of Westminster, 35 Marylebone Road, London NW1 5LS, UK

a r t i c l e i n f o

Available online 12 October 2009

Keywords:

Concessionary fares

Rural

England

Elasticity

User characteristics

Surveys

0X/$ - see front matter & 2009 Elsevier Ltd. A

016/j.tranpol.2009.09.002

esponding author. Tel.: +44 20 7911 5000�3

ail address: whitep1@westminster.ac.uk (P. W

a b s t r a c t

Concessionary fares are offered to those aged 60 and over for bus travel in Britain. This study examines

the effects of replacing a half-fare concession in England by free travel in April 2006. Aggregate data are

used to examine trends; in addition, a large sample was obtained of pass holders in the Salisbury area,

which enables a clear distinction to be drawn between increased travel by those who already held a

half-fare pass, and travel by those taking up the pass for the first time. These data are then used to

derive fare elasticities, and compare the characteristics of new passholders and those previously holding

a half-fare pass. Comparisons are made with studies elsewhere. Overall policy implications are

reviewed.

& 2009 Elsevier Ltd. All rights reserved.

1. The development of free concessionary travel in Britain

1.1. Introduction

Britain is somewhat unusual in offering a very extensive freetravel facility on public transport to a substantial proportion of thepopulation. The largest element of this provision, with which thispaper is principally concerned, is that defined by age group, since2003 those aged 60 upward. A review by Hodson (2008) indicatesthat many European cities and countries offer discounts tocategories such as children and older passengers, with an averagediscount for older passengers of 46% vis a vis the full-priceequivalent (compared with the ‘100% discount’ implicit in freetravel). Concessionary travel is also provided to those experien-cing various forms of disability.

Such provision can be traced back over many years, especiallyin the larger urban areas. The schemes in London, and severalother large urban areas served by Passenger Transport Executives(PTEs), notably West Midlands and Merseyside, date from theearly 1970s. An explicit basis for compensating operators wasestablished under the Transport Act 1985, on the principle thatthey would be neither better off nor worse off as a result of theconcessionary requirement imposed on them by the localauthority concerned. In other words, the operator should notreceive more revenue than that category of passengers wouldhave paid at the applicable fares. Compensation also allows foradditional costs, such as those of extra capacity, although inpractice the great majority is related to the revenue effects.

ll rights reserved.

104; fax: +44 20 7911 5057.

hite).

Hence, where additional trips are generated (or diverted fromother modes) by the concession the operator is not compensatedat the equivalent cash fare for those trips, but a figure is calculatedfor the revenue otherwise received. In practice, the averagecompensation payment per trip is thus considerably less thanthe average fare that would have been paid by the usersthemselves.

Considerable uncertainty exists in such calculations, which arein effect based on estimates of the price elasticity of demand forthis category of passenger. A particular problem arises when freetravel is introduced, since the price reduction is 100%, and issuesarise as to whether existing elasticities can be extrapolated oversuch a range, and the appropriate form of the elasticity function tobe used.

A wide variety of local schemes, including cases where noconcession at all was offered, was found within Britain as a whole,standardisation coming as a result of the Transport Act of 2000,which imposed a common minimum of a half-fare scheme afterthe morning peak for this category of traveller1 from April 2001,with discretion for more generous schemes to be specified by thelocal authority concerned. The 2000 Act also removed powersenabling local authorities to charge for the issue of the pass,making it more attractive to low-frequency users. Within Wales, acommon free fare scheme over the whole of that region wasintroduced in April 2002, and Scotland followed with free fareswithin each area of residence from 30th September 2002.

A much more mixed pattern applied in England until April2006, when a common free scheme – with the same time periods

1 Free travel applies from 0930 to 2300 on Mondays–Fridays, and all day on

Saturdays, Sundays and public holidays.

ARTICLE IN PRESS

S. Baker, P. White / Transport Policy 17 (2010) 20–26 21

as applicable to the half-fare scheme – was introduced undercentral government policy. Subsequently, Scotland adopted theWelsh pattern of permitting free travel through the whole of thatregion for its residents from 1 April 2006, and the English schemehas been modified from 1 April 2008 to permit free travelthroughout England in the same manner as within Wales andwithin Scotland. A review of recent Scottish and Welsh experienceis provided by Rye and Carreno (2008).

The underlying rationale behind such policies has not alwaysbeen explicit. They have not been subject to the detailedevaluation associated with transport infrastructure projects, forexample. In most cases a simple political decision appears to havebeen taken (notably that to extend free travel to all English areasin 2006), possibly influenced by voting power of the older-agegroups. In general, such age groups also tend to have lowerincomes than those in the working-age categories, and henceconcessions directed toward this category could be seen assistingthose on lower incomes (see, however, further discussion of thispoint below).

Transport powers have been devolved to the Scottish Parlia-ment, the Welsh National Assembly and the Northern IrelandExecutive who have all introduced their own independentconcessionary schemes.

In Wales the free concession was intended to address themutual aims of the Welsh Assembly and Local Authorities inrelation to social inclusion, transport and the environment, byreducing financial barriers to public transport for a large part ofthe population, including many people who can be identified aseconomically disadvantaged (Carmarthenshire County Council,2003)

In Scotland the key objectives of the introduction of theconcessionary travel policy are to:

�

Allow older and disabled people (especially those on lowincomes) improved access to services, facilities and socialnetworks by ‘‘free’’ scheduled bus services; and so promotesocial inclusion; � To improve health by promoting a more active lifestyle for the

elderly and disabled;

� To remove the restrictions of the previous local off-peak

concessionary fare scheme, which produced differences inaccess to facilities in different areas of Scotland;

� Promote modal shift from private car to public transport; � Maintain a ‘‘no better, no worse off’’ position for bus operators

with a standard reimbursement rate;

� Provide opportunity for improvements to public transport (e.g.

assist development of multi-operator ticketing; use of im-proved ETM technology);

� Facilitate a more effective administration of the system with

adoption of a standard reimbursement rate (73.6 pence forevery pound of the ‘actual’ fare) and shift of operationalresponsibility from local authorities to Transport Scotland; and

� Provide a stimulus to the introduction of SMARTCARD.

(Halcrow Group, 2009)

This paper is concerned specifically with the impact of freetravel within areas of residence in England from April 2006. Theeffects were often greater in rural regions, in which a half-farescheme was the norm in most cases previously. Due to the need toensure that appropriate compensation is paid to operators, goodquality aggregate data is generally available on the total numberof valid passes on issue, and boardings recorded on buses usingelectronic ticket machines (ETMs) normally employed for cashfare collection, or, to an increasing extent, smart cards. From suchaggregate data, average trip rates per pass may be derived, as well

as total volumes. However, it does not provide a means of readilydistinguishing between travel patterns of existing half-fare passholders whose travel has increased, and those associated withtake-up of passes for the first time. This case study enables sucheffects to be examined in greater depth, by analysing responsesfrom a large sample of pass holders. Unlike the Welsh and Scottishschemes, there is no standard compensation factor, which variesgreatly from one area to another.

Another effect is that of a shift from tokens to passes. A numberof local authorities exercised the option to issue tokens in place ofhalf-fare passes (recipients being offered the choice betweenthem), which entitled users to a certain amount of free travel, byhanding such tokens to the operator in lieu of cash. They couldalso be used for alternative modes, such as taxis. They offered ameans of rationing demand, by imposing an upper limit to theamount of travel made. However, a pass for unlimited travel doesnot impose such a restriction. As such, it is clearly more attractiveto the holder, but increases expenditure incurred by the localauthority in compensating operators. Under the ConcessionaryBus Travel Act 2007 (which introduced the general principle offree travel) local authorities can offer alternative schemes such astokens, but are not obliged to do so (DfT, 2008a, Section 2.3.8).

The issue of distinguishing between travel by existing passholders, and new passholders, in deriving elasticity values isdiscussed extensively by Last (2008), with particular reference totrends in the PTE areas. While an aggregate elasticity value maybe calculated for the whole concessionary category, this tells usrelatively little about behaviour of the travellers themselves andmay be a misleading basis for the extrapolation of the observedchange in travel behaviour to ‘‘full fare’’ trips, required for ‘‘nobetter off and now rose off’’ reimbursement. Where a large growthis found due to a big increase in the number of passes taken up, arelatively large value for price elasticity may be estimated at theaggregate level. However, this may conceal a somewhat smallerelasticity value for those who already held half-fare passes, andthe effect of additional travel by those who did not previously ownpasses. This also has wider policy implications, should the ‘new’passholders display different characteristics to the ‘old’ pas-sholders. For example, if they are more affluent, this might raisequestions about the justification for expenditure on concessionaryfares as a form of reducing social exclusion.

DfT guidance on reimbursing bus operators recommendsdistinguishing as far as practicable between the additional travelby those previously holding a half-fare pass, and that from thosenewly taking-up passes (DfT, 2008a, Section 3.4.18) although theextent to which this is feasible will depend on quality of localdata.

1.2. Overall trends

Data for England as a whole, as shown in the Department forTransport statistics bulletins (DfT, 2008b), shows that total busridership in England outside London and the PTEs rose by 8.3%between 2005/2006 and 2007/2008, following a trend of gradualdecline over previous years. Within the PTE areas there was amarginal growth of 0.9%, also reversing a declining trend, but lessmarked due to the existence of free travel in two of the six areaspreviously. It is likely that the free travel concession extension wasa major factor in explaining this, most areas outside London andthe PTEs having offered a half-fare concession previously.

The DfT data also show that concessionary travel for personsaged 60 and over, and the disabled, rose from 21% to 23% of all bustrips outside London between 2005/2006 and 2006/2007, and‘other concessionary’ from 9% to 11% (DfT, 2007a, Table 36). In theyear 2007/2008, all types of concessionary travel comprised about

ARTICLE IN PRESS

Fig. 1. Map showing the location of Salisbury District. [online] Available from:

http://www.salisbury.gov.uk/living/your-home/moving-house.htm?link=latest

[Accessed: 21 July 2007]. Salisbury District Council.

S. Baker, P. White / Transport Policy 17 (2010) 20–2622

23% of all bus trips in Great Britain, ranging from 14% in London to29% in England outside London and the PTEs, 31% in Scotland and41% in Wales (inferred from DfT, 2008b, Table C). Totalreimbursement for the concessionary scheme in England rosefrom £487 m in 2005/2006 to £712 m in 2006/2007 (DfT, 2007b,Table 1.5).

Extension of pass use to enable free travel throughout Englandsince April 2008 has been reported to have produced a furthergrowth in concessionary travel volumes. Total bus journeys (of alltypes) in the first quarter of 2008/2009 were up by 5.7% on thecorresponding quarter in 2007/2008, the greatest increase beingin English ‘shire’ areas (7.1%) and lowest in Wales and Scotland (at3.2%) (DfT, 2008c). This would be consistent with the wider travelopportunities being a major factor, given that lowest growth wasin those regions where free travel throughout already applied.There is also evidence of increased awareness of the new scheme(DfT, 2008d).

A study by Passenger Focus (2009), a government-funded bodywith responsibility for representing the interests of rail and of bususers, indicates a widespread support of the free concession. Some39% of respondents said that they travelled more by bus withintheir local area since they obtained a concessionary pass. Amongstthose using the pass to travel outside their local area (after April2008), 35% were making journeys by bus that they had previouslymade by car, and 12% making journeys they had not previouslymade by any means, prior to the new concession. General supportfor the concession was reported both from concessionary passholders and other bus users. This is perhaps not surprising,although one must also bear in mind the additional taxationrequired to finance such schemes.

2. The case study area

2.1. Definition

The paper provides a case study of Salisbury District Council(as a Travel Concession Authority, TCA) in southern England,which had offered a half-fare pass before April 2006. A fullerdescription of the work is provided by Baker (2007). In commonwith many such areas, residents were offered a wider area inwhich to use their pass than the statutory minimum, namely thewhole of the shire county in which Salisbury District is located(Wiltshire) and a unitary authority, which was previously a part ofthe same county (Swindon). In addition, the adjoining shirecounties of Hampshire and Dorset were covered (so long as thejourney either started or ended in the home TCA of the passholder). Hence, a wide range of potential destinations for leisureand other purposes became available as well as travel to centreswithin the district of residence. In 2001, 21% of the districtpopulation was aged 60 and over – by 2006, this had risen to 24%.

At the time of the study, Salisbury was district within the two-tier county of Wiltshire (subsequently replaced by a single‘unitary’ authority). A location map is shown as Fig. 1. It couldbe considered broadly typical of rural areas in England, featuring amix of an urban centre, and surrounding lower density areas.

The district had a population at the 2001 census of 114,000,and is dominated by the cathedral city of Salisbury with apopulation of about 41,000, which serves as the main centre forthe area (falling within the ‘medium urban’ size category in theNational Travel Survey, NTS). It also encompasses four smallertowns with over 3000 inhabitants (Amesbury, Bulford, Durringtonand Wilton), in the ‘small urban’ category of NTS. The balance ofthe population lives in settlements of 3000 or under, correspond-ing to the definition of ‘rural’ used in the NTS. Principal busoperator in the area is Wilts and Dorset, part of the Go Ahead

Group. Larger urban areas outside the district include cities withsubstantially bigger shopping centres, and the seaside resort ofBournemouth.

2.2. Increases in pass holding rates and journeys

Total passes issued in Salisbury District to those aged 60 andupward rose by 70.1% between 2005/2006 and 2006/2007. Similaror even higher increases (up to 100%) were observed in otherdistricts within the same scheme. A striking feature of the passtake up rate among those aged 60 upward in Salisbury District,was an increase from 33% of those eligible by age in 2005/2006(and a similar figure in previous years) to 57% in 2006/2007. Theincrease was greatest in the small towns (89%) and rural parts ofthe district (91%), compared with 53% in the city itself. Growth inthe absolute numbers in the age group concerned living in thedistrict was only 1–2% per annum. The increase in take up ratewas greater among those aged 60–69 than in older groups whowere eligible, probably indicating a physical ability to make use ofthe concession. By gender, females had a higher take-up rate thanmales prior to free travel, but their share of total pass holders fellfrom 67% to 60%. Hence, there was a higher rate of increase withineach age category for male pass-holders, including the very oldestgroup (aged 85 upward).

Another feature of the free travel pass was that it became moreattractive to those holding alternative forms of concession, such astaxi vouchers. The net growth in numbers of total pass holders inSalisbury District (including disabled concessions as well as thosedefined by age group) was 62.0%, compared with the 70.1% figuredefined by age quoted above. The absolute number of passes onissue for reasons other than age fell by 62% between 2005/2006and 2006/2007. The great majority of pass-holders are thosedefined by age, but a somewhat different overall average trip ratefor all pass holders might be expected compared with that forthose defined by age as such.

As part of the regular monitoring of ridership for purposes ofcalculating reimbursement payable to operators by the localauthority, data were available on total trips made by all types ofconcessionary pass holder, both within the district studied andthe county as a whole. The total number of journeys made usingall types of concessionary passes in the district rose by 72.5%between 2005/2006 and 2006/2007. Hence, given the net increase

ARTICLE IN PRESS

S. Baker, P. White / Transport Policy 17 (2010) 20–26 23

in the number of all types of pass on issue, the average trip rateper holder rose by about 7%. However, this concealed an increasein the trip rate of those already holding the half-fare pass, and theoffsetting effect of the lower trip rate among those taking up apass for the first time, together with possible complicationsarising from pass holders defined by categories other than age.National evidence suggests an average growth in concessionarypatronage (outside London and the PTEs) of 73% between 2005/2006 and 2006/2007 (DfT, 2008a, Section 3.4.42), suggesting thatthe outcome in Salisbury is typical in this respect.

2.3. The sample survey

It was possible to carry out a direct survey of pass holders(defined by age eligibility) renewing their passes in Spring 2007, apostal self-completion form providing data on topics such as tripscurrently made, and changes since the free concession hadreplaced the half-fare pass. It was possible to distinguish betweenthose who had previously held a half-fare pass, and those who hadnot taken up the concession until travel became free (thisincluded not only those reaching the eligible age during thatperiod, but also those already of 60 and over who had notpreviously taken up the opportunity of holding a pass). Thesample could also be sub-divided by area of residence, betweenthose living in the city, the smaller urban areas, and the ruralareas of the district.

Some 796 useable questionnaires were returned, representinga very good response rate of about 61%. Of these, 48% wereresidents of the city, 20% of the smaller urban areas and 32% of therural parts of the district, compared with proportions of bus passholders by area at that time of 48%, 15% and 38%, respectively,suggesting broadly representative sample, albeit some over-representation of smaller urban areas, and under-representationof rural areas. Some 51.8% of the sample had previously held ahalf-fare pass, a slightly smaller percentage than found amongtotal passes then on issue, of which 58.7% had formerly been heldas half-fare passes. It is possible that those benefiting from freetravel for the first time were more likely to respond to a survey.This would tend to produce a slightly lower growth in overall triprate vis-a-vis the average trip rate for half-fare pass holders priorto the introduction of free travel, than would be derived fromaggregate data. However, the trip for rate for new pass holders assuch could be biased upward if those making greater use of theconcession were more likely to respond.

Respondents were invited to estimate their current weekly tripfrequency by selecting a category, from which an average trip rateper week for the sample was estimated. For those previouslyholding a half-fare pass, their trip frequency before the introduc-tion of free travel in April 2006 was also requested. About 80% ofrespondents reported at least one trip per week (for the wholesample, after free fares were introduced), broadly similar to thesample obtained in the Passenger Focus study (Passenger Focus,2009, Fig. 3-1).

The overall weighted average trip rate thus estimated for 2006/2007 from the sample was considerably higher than the averagetrip rate from aggregate data for 2006/2007 referred to above.This suggests those responding to the survey have a higher triprate than holders in general, which is not wholly unexpected,since pass holders benefitting from a higher trip rate might bemore likely to respond to a self-completion survey, and the surveyexcluded disabled holders who may have a lower rate. The mainvalue of the survey lies in comparing relative differences betweenhalf-fare pass holders and new pass holders after free travel wasintroduced, together with differences by type of area. Amongthose already holding a half-fare pass, there was evidence of a

greater increase by those living in the rural areas, where the costper trip (even at half-fare rates) would have been highest. Amongthe new pass holders, there was very little difference by area.

Total percentage growth in total concessionary pass holder tripvolumes inferred from the survey sample was noticeably higher inthe case of rural areas, associated with greater growth in pass takeup.

A substantial proportion of those previously holding a half-farepass (about 39%) stated that their level of bus pass use wasunchanged with the free pass, and 2% reported a decline (whichcould be due to other personal factors). However, 59% ofrespondents stated that their level of bus pass use had increased,with 28% stating that their usage had either ‘doubled’ or ‘morethan doubled’, 10% increased by a quarter and 22% increased ‘alittle’. Given the rather generalised nature of this question, inestimating changes in trip rates per week (as shown below), anaverage was derived from reported trips in terms of frequencycategories. This produced a somewhat lower figure (an averagegrowth of about 17%), which in turn affects elasticity estimatesshown.

Sensitivity tests were carried out on the trip rates assumed fora stated frequency category (for example ‘4 or more days perweek’), but these had very little effect on the overall ratios of‘before’ and ‘after’ weekly trip rates for given user-frequencycategory, provided that the same rate was applied in the ‘before’and ‘after’ cases’. It could be argued that the upper statedfrequency of ‘4 or more days per week’ could conceal some veryhigh trip rates, which might have increased in the ‘after’ period,but even assuming an average of 10 single trips per week (i.e. twosingle journeys on 5 days) this had little effect. It is unlikely thatmany former half-fare users would have travelled as often as 6 or7 days per week even at free fares, and hence an upper estimatefor the average of 10 trips per week would be reasonable.

2.3.1. Relative trip rates derived from Salisbury District pass-holder

sample

In comparing the reported weekly trip rates from the survey,these are shown below relative to the trip rates for those holdinghalf-fare passes prior to the introduction of free travel. Theweighted average for the whole area covered is shown as a valueof 1.00.

Trip rate by holders of half-fare passes, prior to free travel

1.00 Variations by area of residence Salisbury city 1.19

Small urban

0.92 Rural 0.71

As might be expected, a higher trip rate was found in Salisburycity itself, given the higher availability of bus services.

On the same base, the reported trip rates of those who heldhalf-fare passes, after introduction of free travel were as follows:

Overall average

1.17 Variations by area of residence Salisbury city 1.39

Small urban

1.04 Rural 0.87

Reported trip rates of new bus pass holders, after introductionof free travel:

Overall average

0.86 Variations by area of residence Salisbury city 1.02

Small urban

0.83 Rural 0.68

ARTICLE IN PRESS

S. Baker, P. White / Transport Policy 17 (2010) 20–2624

These are below those of those previously holding half-fare

passes, which may be associated with factors such as higher carownership (discussed further below), and possibly some bias insurvey response toward those with higher trip rates. The relativedifferences between Salisbury city and rural trip rates are similarto those of other groups (the rural trip rate being 67% of theSalisbury city figure).

The overall average trip rate ratio to that by all pass-holders,after free travel, was 1.02.

The implied overall average trip rate increase of about 2% isslightly lower that from aggregate data of about 7% (as citedabove), suggesting a broadly representative picture from thesample, despite higher absolute trip rates.

Among those who previously held a half-fare pass, the overallaverage increase in weekly trip rate was about 17%, which wasvery similar within Salisbury city itself, but noticeably lower insmall urban areas (about 13%), and higher in rural areas (21%).

The weekly trip rates of those newly holding passes were some73% of those of previously holding the half-fare pass. This isconsiderably greater than the value of about 50% suggested in DfTguidance, based on Welsh experience (DfT, 2008a, Section 3.4.17).

An interesting implication of such trip rates is the proportionof net absolute growth accounted for by new pass holders.Whether using the sample-weighted base, or a population-weighted base, this may be derived using the trip rates above,and indicates that about 80% of the net growth was from thissource. This has important revenue implications for bus operatorssince, while some of the new pass holders might previously haveheld an alternative form of concession (such as tokens), those whopreviously paid their own fares (albeit at a lower trip rate) wouldhave paid the full adult fare to the operator.

A limitation of the survey is that the ‘new’ passholders werenot asked to estimate their trip rate prior to holding the pass,which could have enabled more explicit conclusions to be drawnabout the extent of their bus travel by other means of paymentprior to holding the pass. Ideally, a panel survey of those holding ahalf-fare pass prior to April 2006 and free pass thereafter wouldalso have enabled a more precise comparison of trip rates thanusing recollection, but it was only feasible to conduct the surveyat one point in time. It would not, in any case, have been possibleto identify the ‘new’ passholders for a panel survey prior to theirtaking up the concession.

2.4. Differences between old and new pass holders

As indicated above, the ‘old’ pass holders (those previouslyholding a half-fare pass) differ substantially from the ‘new’ passholders (those taking up a new pass after free travel came intoeffect). The former displays a higher trip rate than the latter.Analysis of the questionnaire data revealed further importantdifferences, all of which were found to be highly significant on aw2 test:

�

‘old’ pass holders were far more likely to indicate that their‘main mode’ of transport was the bus – 63.5%, compared with31.1% for the ‘new’ pass holders. Conversely, only 33.6% of ‘old’pass holders indicated that car (as driver or passenger) wastheir main mode of transport, compared with 64.7% for ‘new’pass holders. � in addition to differences in the average trip rate per week, a

different distribution was also found. Whereas 27.9% of ‘old’pass holders used their pass on four or more days per week,this applied to only 14.3% of ‘new’ pass holders. The latterhad higher occurrences of ‘1 day per week’, and lower,frequencies.

�

‘old’ pass holders were far more likely to live in a householdwithout a car – 70.3% – than did the ‘new’ holders. � ‘new’ pass holders were more likely to possess a driving

licence.

� ‘old’ pass holders were more likely to be in receipt of council

tax or housing benefits than ‘new’ pass holders.

� ‘old’ pass holders are also likely to be found among the oldest

age groups, forming a successively higher proportion of all passusers sampled as higher age groups are considered, in additionto the obviously higher proportion of ‘new’ pass holders amongthe 60–65 age group.

Hence, the ‘old’ pass holders are more likely to be the low-income, low car ownership, bus-dependent type of person that onmight traditionally associate with concessionary fare bus use.

This evidence points to a greater degree of discretionary use ofbuses by the ‘new’ pass holders. They might, for example, retainuse of the car for most trips they make, but be likely to divert tobus for journeys into town centres where congestion and/orparking charges affect the relative attractiveness of using a car.Some parallels may be drawn with the Passenger Focus study,which indicates some modal shift from car to bus associated withthe free bus pass, mainly in respect of trips to town and citycentres with parking space and congestion constraints, althoughfor those pass holders who did have access to a car it was still thedominant mode for other journey purposes less accessible by bus(Passenger Focus, 2009, Sections 3.17–3.19).

The policy conclusion may be that outcomes of the largeincrease in pass holding is less closely related to any socialexclusion objectives than the original aim of concessionary travel,but could be aiding other transport objectives, such as achieving ashift from car to bus.

2.5. Price elasticities

From data on estimated trip rates and volumes, it is possible tocalculate price elasticity values. These are also dependent uponthe form of elasticity definition adopted (for a fuller discussion,see Balcombe et al., 2004, Chapter 5).

As a simple starting point, the ratio of demand change to realfares change can be calculated. For those already holding the half-fare pass, the monetary fares change was minus 100%, and averagegrowth in reported trip rate 17%, giving an elasticity of �0.17. Thisis somewhat larger for rural areas (�0.22) and smaller for smallurban areas (�0.13). However, any shift from a paid fare to freetravel is, by definition, a reduction of 100%, irrespective of theabsolute base. In rural areas, the absolute average fare would besomewhat greater, and hence a larger elasticity might thus beexpected.

In respect of the entire market, a much larger magnitudewould be derived, since the 100% reduction was associated withan aggregate demand growth of trips on all types of concessionarypass of 72.5%, a crude elasticity ratio of �0.72. However, thiswould be complicated by shifts between different types of passand alternative concessions. More importantly, it implicitlyassumes the concessionary travel by ‘new’ passholders corre-sponded entirely to a net growth in bus use, whereas in practice itis likely that some of this travel would have been made at normalbus fares. Hence, the elasticity values would be exaggerated.

There are also problems in comparisons with establishedelasticity values where free travel is introduced, since themonetary cost of a trip falls to zero, a somewhat extreme caseover which to extrapolate values from earlier research. However,the total ‘cost’ to the user in a wider sense does not become zero –there is still the element of travel time and inconvenience

ARTICLE IN PRESS

S. Baker, P. White / Transport Policy 17 (2010) 20–26 25

involved in a journey. Using a method similar to that adopted inthe study for the Welsh Assembly Government (MVA consultancy,2003), estimates were made for fare elasticities for equivalentchanges of 1% and 10% at the full adult fare, based on an assumedgeneralised cost (prior to effect of free travel) composed of anaverage fare of £1 and journey duration of 30 min in-vehicle, plus10 min waiting time and 6 min walking time (a generalised timeof 62 min) – with variations by area type – and value of time of£5.87/h. For ‘old’ pass holders the resultant elasticity was about�0.27 (using the sample data), ranging from about �0.26 inSalisbury city to �0.35 in rural areas.

DfT guidance on concessionary fares compensation is based ona point elasticity at the full-fare level, assuming a negativeexponential demand function (DfT, 2008a, Section 3.4.8) for themove from half- to free-fare travel, as follows:

Elasticity at full fare¼ � lnðq2=q1Þ=ðf1 � f2Þff

where q1 is the demand at the half-fare concession level, q2 thedemand after introduction of free travel, by those previouslyholding half-fare passes, (in this case, the ratio q2/q1=1.17, asabove), f1 the fare at full-fare equivalent level (assumed at £1, asabove), f2 the fare at half-fare level (i.e. 50 pence), ff=full-farelevel (i.e. £1).

Applying data from the Salisbury case gives an elasticity of�0.31. DfT guidance (DfT, 2008a, Table 2) suggests a centralestimate of �0.5 for ‘other urban’ (range �0.30 to �0.75), and of�0.75 (range �0.50 to �0.90) for ‘rural’.

2.6. Types of journey made

Salisbury was stated by 80% of respondents as their mostfrequent destination when using the bus. This can be explained bythe dominance of the centre as the main service and retail centrein a mainly rural district, along with it acting as the hub in the busnetwork.

The most frequent journey purpose for bus trips made with thefree pass was food shopping at 42%. Shopping of any typerepresented the main purpose of the vast majority of conces-sionary bus trips. This is consistent with research on conces-sionary pass holders in Scotland.

3. Overall impacts and conclusions

Overall bus use in Wiltshire, within which the case studydistrict is located, rose by 19% between 2005/2006 and 2006/2007, considerably more than the average for areas outsideLondon and the PTEs. While some of this was due to expansionof park & ride use, and specific bus service quality improvements,it is likely that the majority was a result of the free concessionarytravel (concessionary trips rose from 20% to 27% of all trips withinthe county over that period).

The high volume of bus use by those previously not holding apass is a noteworthy feature. While their trip rate remains fairlylow, it has increased the total bus market substantially. Hence, thebus has become of importance to a greater number of people thanbefore. Those in the rural areas, where fares were previouslyhigher, may have received the greatest benefit. There appears be agreater use of buses by those with cars available. This in turn maysupport wider policies aimed at traffic reduction in such cases, aswell as the more traditional policy objective of aiding mobility bythose without access to cars. However, a large increase in publicexpenditure has occurred as a result and all those aged 60 or overbenefit, irrespective of their income.

Implications for operators are somewhat more complex,especially insofar as compensation is based only on the change

in behaviour of those previously holding a half-fare pass wouldunderstate the revenue loss due to some trips made by the ‘new’pass holder having been at full adult fares.

To what extent can findings be extrapolated from this casestudy to other rural regions? There has been relatively littlecomprehensive monitoring of the impacts of extending freeconcessionary travel throughout England, despite the large publicexpenditure thereby incurred and numerous disputes regardingcompensation due to operators. As shown earlier, the overallincrease in concessionary travel appears close to the average forareas outside London and the PTEs. A limitation of the study isreliance on recall of trip rates for those previously holdinghalf-fare passes, and lack of equivalent data for the new passholders. Hence, estimates of changes in trip rates are less precisethan one would wish, and the elasticities quoted should beregarded as only approximate. However, much firmer conclusionsmay be drawn in respect of the characteristics of persons formerlyholding half-fare passes, and the ‘new’ passholders.

Clearly, it should be possible in all other areas to estimate theincrease in passes on issue and total trips attributed to them, thusinferring changes in the average trip rate. Where a large increasein total passes on issue has occurred but only a small changein overall average trip rate per passenger, it would be reasonableto assume that an outcome similar to that described in this casestudy has occurred, i.e. growth in trip rates would be found forthose already holding a half-fare pass as a result of free travel,while new users display lower trip rates.

Where smartcard technology is in use, far more extensiveopportunities for data analysis are presented (Bagchi and White,2005). Variations in trip rates can be identified as well as changesin overall averages. While retaining anonymity of individual cardholders, it should be possible to separate samples of cards heldby those previously holding a half-fare pass from those newlytaken-up after April 2006 (although over time this distinctionwill become less valid, due to underlying turnover in the market).Hence, a trip rate for each category can be identified, as well astimings of trips by time of day and day of week. Where furtherdirect survey work is undertaken, it would be desirable to identifycar ownership levels among pass holders, and modes previouslyused for trips that have shifted to bus, especially by new holdersof passes after April 2006. This would enable testing whether thefindings described above – such as that new, lower-frequency passusers are more likely to have higher car availability – also applyelsewhere. Further impacts since the extension of free travelthroughout England in April 2008 can be likewise assessed.

Acknowledgements

The Wiltshire and Swindon concessionary schemes areadministered on behalf of the local authorities concerned by theconsultancy MCL, to whom we are grateful for availability ofaggregate data.

Helpfulness of the respondents to the survey is muchappreciated.

At the time of undertaking the research, Stuart Baker wasemployed by Salisbury District Council, whose support isacknowledged.

All conclusions drawn from data are those of the authors alone.

References

Bagchi, M., White, P.R., 2005. The potential of public transport smartcard data.Transport Policy 12 (5), 464–474.

Baker, Stuart (2007). Standing Room only?: The impact on bus use resultingfrom the move from a half-fare to a free fare concessionary bus pass in a shirearea of England—case study of Salisbury District, Wiltshire. M.Sc.

ARTICLE IN PRESS

S. Baker, P. White / Transport Policy 17 (2010) 20–2626

Research Dissertation, Transport Planning and Management, University ofWestminster, 2007 (unpublished).

Balcombe, R., Mackett, R., Paulley, N., Preston, J., Shires, J., Titheridge, H., Wardman,M., and White, P. (2004). The Demand for Public Transport: A Practical Guide.Transport Research Laboratory Report TRL593.

Carmarthenshire County Council, 2003. Environment and InfrastructureScrutiny. Concessionary Fares in Carmarthenshire’. Accessed at http://online.carmarthenshire.gov.uk/agendas/eng/SEIN20031125/SUM05. HTM 24 August2009.

Department for Transport 2007a. Public Transport Statistics Bulletin GB: 2007Edition Supplement. Department for Transport, November 2007 (publishedonly in electronic form, via website www.dft.gov.uk/transtat/).

Department for Transport 2008a. Concessionary travel for older and disabledpeople: guidance on reimbursing bus operators.

Department for Transport 2008b. Public Transport Statistics Bulletin GB: 2008Edition. Department for Transport Statistics Bulletin SB(08)23, September2008, Table C.

Department for Transport 2008c. Bus and Light Rail Statistics GB: April–June 2008.Transport Statistics Bulletin SB(08)10, September, Table 3.1.

Department for Transport 2008d. Awareness and use of bus concessionary fares.[Report of survey by ONS].

Halcrow Group Ltd. 2009. The Scottish Government. Transport ResearchSeries: Evaluation of Concessionary Travel in Scotland. Accessed at http://www.scotland.gov.uk/Publications/2009/05/13144419/0 24 August 2009.

Hodson, P. (2008). Social fares in urban public transport in Europe, PublicTransport International [UITP] November/December, pp. 17–19.

Last, A. 2008 Free Travel for Older and Disabled Passengers—Lessons for TransportPlanners. In: Proceedings of Sixth Transport Practitioners Meeting, Universityof Reading, July (available on CD Rom from PTRC Ltd.).

MVA consultancy 2003. All-Wales Concessionary Fares Reimbursement Study—FinalReport.

Passenger Focus 2009. England-wide Concessionary Bus Travel: the PassengerPerspective. Report for Passenger Focus by Integrated Transport Planning Ltd., July.

Rye, T., Carreno, M., 2008. Concessionary fares and bus operator reimbursementin Scotland and Wales: no better or now worse off?. Transport Reviews 15 (4),242–250.