Benefits to their communities from small town professional football clubs.
Barlow, Andy ; Forrest, David
Many small town professional clubs operate on the border of
viability and threats to their survival are often real. But their
activities may generate benefits that they are unable to capture in
ticket sales; for example they provide a focus of interest even for some
residents who do not attend the stadium. The paper presents results from
applying contingent valuation methodology in two towns with clubs in the
bottom-tier of English professional football. Collective
willingness-to-pay for the survival of the clubs through taxation proved
to be significant relative to the revenues the clubs were able to
generate through ticket sales. There appears therefore to be a case for
local governments to consider intervention in some circumstances where a
club's survival is in question.
Keywords: football; contingent valuation; willingness-to-pay
JEL Classifications: D62, H41, L83
Introduction
Although most attention, and revenue, in football accrues to clubs
based in major cities and playing in premier leagues, the majority of
professional clubs in Europe play in smaller centres and in lower
divisions. Many such clubs operate on the border between viability and
non-viability (Szymanski, 2014). Sometimes, their survival may depend on
explicit or implicit public subsidies (for example, through relief from
local property taxes or, as is common in Spanish football, provision of
a rent-free municipal stadium) or other public policy interventions (for
example, protection through planning laws against acquisition of stadium
sites by property developers). This paper asks whether civic authorities
should in fact be concerned over whether their communities are
represented in their national professional football league. If a club
were to collapse or else forced to become amateur, would that be a
legitimate matter for concern from a local public policy perspective? If
the revenue of a club is insufficient to cover the costs associated with
playing in the professional league, is it ever appropriate for local
government to override the 'market solution' ? Is it that
there are additional benefits to a community from the existence of its
professional football club that are not captured by the revenue the club
collects at the stadium? Is it possible to measure these community
benefits which lack a corresponding revenue stream? Are they potentially
large enough to justify intervention if ticket revenue fails to cover
the costs necessary for the club to maintain its professional status?
This paper addresses questions such as these through investigation
of two examples of clubs in the fourth (i.e. the lowest) tier of the
English professional league structure. The English game offers one of
the largest professional structures in the world, organised as four
divisions, of which the top is the English Premier League, all linked
with each other by a system of promotion and relegation. In principle,
of course, any member club could slip down into the bottom tier. But, in
practice, it consists largely of small town clubs, consistent with the
finding in Buraimo et al. (2007) that the position of a club within the
hierarchy of the 92 English professional league clubs is determined
largely by size of local population.
It is true that small-town clubs demonstrate impressive resilience.
We looked at all of the clubs which played in the Football League eighty
years ago and, while several subsequently failed or were demoted from
the League, only three were not reformed and all the rest are still
competing at some level of the game today. (1) In the modern era,
Szymanski (2014) notes that there were 67 cases of insolvency during
1984-2010 but all the clubs present in the League at the beginning of
the period were still active at the end albeit eight of them were
outside the professional structure.
So, history suggests that clubs seldom die but that they do drop
out of the professional league into the amateur and semi-professional
structure, either because they experience financial collapse (as with
Maidstone United in 1992) (2) or because their weak finances lead to
inadequate playing performance and demotion to the minor leagues (for
example, towns such as Barrow and Southport proved incapable of
sustaining a club at the professional level). (3) From a local point of
view, our interest is in whether civic authorities should proactively
seek to maintain their professional club. Could the inability of a
professional club to survive be an example of 'market failure'
and, if so, is the welfare loss large enough for concern? From an
English national perspective, the relevance of our research is to the
question of whether four national, fully professional divisions are too
many. Could the smaller centres be adequately served by amateur
football, dispensing with the need for the fully professional game in
England to sustain as many as 92 clubs and four divisions?
In the next section, we draw on the framework of welfare economics
to examine whether and how it would be possible to justify public
intervention to ensure that a town keeps its professional football club.
We argue that there is a potential case because it is plausible that
clubs generate benefits to the community which are not captured in their
revenue streams. But the quantification of these benefits is necessary
to reveal the strength of the case for offering a club support.
Therefore, we carried out surveys in two relatively small towns where
local fourth-tier clubs were under threat, employing the contingent
valuation method (CVM) to evaluate local willingness-to-pay to sustain
professional football in the community. Following a review of previous
applications of CVM in the sports sphere, we will describe our surveys
and the contexts in which they were conducted. Then we will present
results from the two surveys in terms of estimates of community
aggregate willingness-to-pay for professional football to continue to be
offered in each town. The final section of the paper will pull together
the principal conclusions from the whole exercise.
Sources of market failure
Standard welfare economics identifies two scenarios relevant to the
present context where it may be socially efficient for an enterprise to
continue to exist even though the operator of the enterprise cannot
generate sufficient revenue for supply costs to be covered at any level
of unit sales. In such circumstances, a case may be made for
intervention, for example through subsidies or through a change in the
mode of ownership (such as to municipal ownership, with trading losses
then met from local taxes).
The first of these two categories of 'market failure' can
arise where the provider of a good or service possesses a degree of
market power such that the demand curve facing the firm is downward
sloping. This is plausible in our context because football clubs will be
price-makers rather than price-takers, given that many attendees would
regard travelling to watch a club representing a different town as a
very imperfect substitute. It is also plausible, as in other industries
offering public performances as their product, that fixed costs
(dominated, in this case, by the wage expenditure required to hire
players good enough to maintain membership of the league) (4) account
for such a high proportion of total costs that the average cost curve is
also downward sloping throughout the relevant range of demand. With this
scenario, the downward sloping demand curve may lie entirely inside the
downward sloping average cost curve and there would be no level of
ticket price at which costs would be covered. The football club would
then be commercially non-viable and the market would not supply
professional football in that community. However, the aggregate
consumption benefit to users of the product (attendees) will be given by
the area under the demand curve, (5) and there may then be a ticket
price/level of attendance where aggregate consumption benefit would
exceed total cost even though total club revenue would fall short of
total cost (figure 1). The possibility arises because the club will
price tickets to maximise revenue (minus variable costs) and price has
to be set according to the strength of preference for the product among
marginal attendees. Further up the demand curve, there will be greater
willingness-to-pay, for example among hard core supporters, but the high
value of the satisfaction these users obtain from the product is not
captured by the club in revenue but remains as 'consumer
surplus'. If the value of consumer surplus is high enough, and if
the club cannot extend price discrimination to capture it, it may be
socially efficient to intervene to secure continued provision of the
service. (6)
If this argument were indeed to be employed to justify intervention
in the case of football, it would be necessary first to evaluate annual
consumer surplus. Alexander et al. (2000) provided impressively high
estimates of levels of consumer surplus in American major league sports
and argued that consumer surplus must therefore be taken seriously if a
subsidy is proposed to maintain a franchise. But their estimates used
only revenue data and are sensitive to the value of attendance demand
elasticity assumed. The travel cost method provides a means for
estimating demand for sports (Forrest et al., 2002, Hakes et al., 2011)
and therefore of measuring consumer surplus more reliably; but, in the
case of bottom-tier clubs, the catchment area is liable to be tight, and
there will then be insufficient variation in the origins of trips to the
match to permit a well-defined demand function to emerge. The only other
alternative appears to be to measure consumer surplus by a stated rather
than a revealed preference method. If an individual user of the service
truly values it at more than he has to pay for it at the ticket office,
then he would be willing to pay a lump sum per year to secure the right
to choose to attend games at current prices. In principle, consumer
surplus could therefore be measured (so long as honest answers could be
elicited) in a survey which asked attendees about the maximum they would
be willing to pay to prevent the product from being withdrawn from sale
(professional football no longer available in the town).
[FIGURE 1 OMITTED]
A second category of market failure identified in welfare economics
is that of externalities. Here again, clubs may generate benefits but be
unable to capture corresponding revenue, biasing the market against the
provision of the product: if (positive) externalities have a
sufficiently high value, they may then justify intervention to prevent
the loss of professional football in a town. The notion that such
externalities exist for sport, and may well have high aggregate value to
the local population, was proposed by Noll and Zimbalist (1997). For
example, a significant proportion of the residents of a town may gain
consumption benefit because they identify with the local club and it
provides interest when they follow the club's progress in the
press. Similarly, they may engage in more social interaction because the
results of the local club are a topic for conversation; or they may
simply enjoy hearing the name of their town mentioned in the national
media when the football results are read out. (7) Benefits such as these
may be termed as comprising a 'public good' (a public good is
an extreme case of externalities where the benefit is available
automatically to everyone and no one can be prevented from taking
advantage of it). In principle, the annual value of the public good
could again be assessed through a survey which asked residents about the
maximum they would be willing to pay to prevent professional football
from being lost to the town. (8)
The relevance of both potential sources of market failure can
therefore be assessed by asking the same question of users and non-users
(those who attend and those who do not). In the case of non-users (who
never attend the stadium), responses will reflect their valuation of the
public good the football club produces. In the case of users (those who
attend matches), their responses should aggregate the value of consumer
surplus associated with use of the private good (excess of consumption
benefit from attendance over amount spent on tickets) and the valuation
they place on the public good, which can be consumed independently of
attendance. (9) From a social perspective, intervention is justified if
the annual subsidy required to maintain football in the town is less
than the sum of the annual value of the consumer surplus that would be
lost by attendees if it ceased to exist and the annual value of public
goods associated with the club (where these accrue to both attendees and
non-attendees). Answers to a properly framed survey question would
reveal the sum of these two components, which is what we require to
know.
Contingent valuation
The notion that the value of positive externalities could be
assessed simply by asking a sample of the population what they would be
willing to pay to secure them was first proposed, in the context of soil
conservation programmes, by Ciriacy-Wantrup (1947). When this approach,
termed subsequently the contingent valuation method (CVM), was put into
empirical practice, it was again typically in the context of
environmental policy. By the 1990s, environmental economists had
sufficiently refined the methodology that the US courts judged it
reliable enough to be the basis for calculating damages payable by those
responsible for a major oil spillage where public amenity had been lost.
The requisition of CVM to the service of sports policy is more
recent but appears to be becoming commonplace. Atkinson et al. (2008)
and Walton et al. (2008) assessed the value placed by UK residents on
hosting the 2012 Olympic Games. Relatively few would attend events; but
there was still an indication of high anticipated non-use value.
Similarly, Preuss and Werkmann (2011) estimated that the German
population had an aggregate willingness-to-pay to host the Winter
Olympics of more than 0.9bn [euro]. Wicker et al. (2012a and 2012b) used
CVM surveys to value the benefits of national (German) success in the
Olympic Games and football World Cup respectively. Again substantial
benefit was identified (for example, a gold medal in athletics had a
mean valuation of more than 5 [euro] among German adults). In all these
cases, the tangible benefits, be they of hosting or winning, appear low
relative to the investments involved; but CVM reveals additional and
decidedly non-trivial benefits from the 'public goods'
generated in terms of public satisfaction; and this provides a basis for
believing that the expenditure of public funds might be justified after
all. Generally, these results validate the recommendations by Baade
(2006) and Walker and Mondello (2007) that evaluation of sports policy
should move away from an emphasis on economic impact studies (where,
when properly conducted, benefits are usually shown to be very low) and
towards the use of CVM, to estimate the consumption value associated
with the public goods that sports produce.
Our particular concern is with the value associated with
availability of local professional sport. Here the bulk of research has
been in American settings, where the threat of losing a professional
club may be very real given that the owners are free to move their
franchise to another city. For example, Johnson et al. (2001 and 2007)
inquired into willingness-to-pay to maintain a National Hockey League
franchise in Pittsburgh and into willingness-to-pay to keep a National
Football League club in Jacksonville. Fenn and Crooker (2009) considered
the case of Minnesota's team in the National Football League. In
each case, aggregate willingness-to-pay was substantial, but not as high
as the cost of a new stadium that owners might require the city to build
to prevent a move.
In Europe, Castellanos et al. (2011) took a similar approach to the
American CVM studies when they estimated the public's
willingness-to-pay to keep a Spanish football club, Deportivo de A
Coruna, in the First Division. Note that, whereas the American model of
sport permits migration of franchises such that any threat of removal
involves the professional club completely ceasing to exist in the city,
such an extreme scenario is implausible in the European model.
Castellanos et al. therefore asked about willingness to make a financial
contribution to ensure "that Deportivo could maintain the
competitive level of recent seasons". This could be interpreted as
defining the good to be valued as a club performing at a level
sufficient to maintain its (then, 2003) current First Division status.
Similarly, to create a plausible scenario, we will address valuation of
the existence of English local clubs performing to a level consistent
with maintaining professional league (fourth-tier) status. In A Coruna,
mean annual willingness-to-pay by respondents to the survey was
approximately 10 [euro].
The previous literature reporting CVM research on local sports
teams has focused on the highest levels of their respective sports, for
example the National Football and Hockey Leagues or the top tier of
Spanish football. Here, an innovation is that we move the focus closer
to the grassroots by evaluating benefits from small-time clubs in
small-town communities. We had no priors on the size of per-person
benefits compared with those generated by major league clubs. Certainly,
the top-tier clubs attract higher volumes of media attention and there
is therefore 'more public good' to be consumed. On the other
hand, small towns have fewer local amenities and a club that has low
status within its sport may nevertheless be an important focus for civic
pride and public attention and may promote feelings of identity. In this
sense, it may contribute importantly to local social capital. We were
unable to rule out a priori that, at least per person, local residents
would value a 'minnow' of the football world just as highly as
local residents in bigger cities valued their much more celebrated
clubs, playing at the tops of their sports.
Methodology
Any stated preference approach runs the risk of eliciting responses
that are inaccurate, for example because the respondents do not have
enough information to express an informed choice or because the question
put is too hypothetical or because they engage in strategic behaviour.
(10) As CVM has evolved, it has developed survey techniques to minimise
the risk of eliciting misleading responses and we attempted to reflect
current best practice in the decisions we describe in this section on
how the exercise was to be conducted. Broadly, our survey was similar to
those employed by Johnson et al. (2001) and Castellanos et al. (2011).
Castellanos et al. (2011) argued the merits of adopting similar
questionnaires to facilitate comparison between the valuations of
professional sports in different settings.
The first potential pitfall faced in conducting a CVM exercise is
that respondents may overstate their true willingness-to-pay if faced
with a hypothetical choice (Loomis, 2011), particularly if it is outside
their realm of experience and difficult to take seriously. The risk of
'hypothetical bias' is reduced if respondents are invited to
consider a realistic scenario. With this in mind, we selected the towns
of Bury and Luton as the settings for our investigations.
Bury is a town on the northern edge of Greater Manchester County,
about 13 km. from the central city. Its social composition mirrors that
of the UK to the extent that its two parliamentary constituencies are
regarded as 'bell weather': which party wins is a good
predictor of the national result. Luton is a manufacturing town, 50 km.
north of London. Each had a population, defined by the boundaries of the
municipal government, of approximately 185,000, with 70,00080,000
households. Luton's population exhibits much the greater ethnic
diversity (table 1).
In Bury, the survey was administered in April, 2008. The
town's club, Bury F.C., had ridden uncharacteristically high in the
division standings during the Autumn; but a long slump in results had
taken it to 21st out of 24 clubs and its form suggested that it was in
real and imminent danger of finishing in the bottom two places in the
league and thereby losing the professional status it had maintained
since 1885. The situation for Luton Town F.C. was similar when we
applied the questionnaire in the town during 2009. (11,12)
In each town, the scenario we put to respondents was therefore
relevant to the then current situation of the local football club.
Respondents were reminded about the current dangers facing the club and
then asked what would be the highest monthly amount they would be
willing to pay to ensure the club's survival in the professional
league. The question was to be answered by selecting from preset
amounts, at 2 [pounds sterling] intervals, between zero and 20 [pounds
sterling]. (13) Such closed-ended, as opposed to open-ended, value
elicitation questions mitigate the risk of hypothetical bias (according
to meta-analysis of CVM environmental studies by Murphy et al., 2005).
Potentially they may also remove a barrier to answering the question at
all if respondents would be daunted by an unfamiliar question posed
without any prompt about the range of values they should consider. (14)
Together with a scenario (description of the amenity to be provided
or protected) for the respondent to evaluate, it is important also to
specify a plausible payment mechanism. The questionnaire should
discourage strategic responses whereby an individual exaggerates his
willingness-to-pay because he believes it would in fact be others who
would fund the amenity under consideration. To combat this risk, the
payment vehicle should therefore be plausible such that the respondents
should believe that they would actually have to pay in the event that
the project (in this case saving the football club) was approved. In
their Spanish study similar to ours, Castellanos et al. (2011) asked
about willingness to make voluntary contributions into a fund to support
the football club. This might elicit dishonest answers because of
strategic response behaviour (which would lead to overestimation of the
value of the club); but even 'honest' answers may be
misleading if there are respondents who value the amenity but who would
expect to be able to free-ride because enough others would contribute
(stated willingness-to-contribute may then underestimate the value
placed on the amenity).
We chose council tax as the payment vehicle. This is a residential
property tax levied by the town council and chargeable to the large
majority of households (albeit there is a 25 per cent discount for
single occupancy and a system of reimbursement for those on welfare
benefits). Council tax was selected as the payment vehicle primarily
because it would be a believable part of the scenario offered to
respondents. But it has the additional merit that nearly every resident
would be familiar with the fact that it is used to fund locally
organised services such as libraries. This should encourage respondents
to think in terms of the opportunity cost of expenditure out of council
tax revenue.
That council tax, which most households in the towns paid in
monthly instalments (hence our focus on monthly willingness-to-pay),
might be used to guarantee the survival of the local club was likely to
be especially believable in Bury and Luton because the councils already
provided limited funding for the clubs. This is relatively unusual in
England and was a further reason for our choice of the two towns for the
surveys. In Bury, support was through sponsorship of the team shirt and
in Luton the club was allowed a discount in the amount of business rates
(local tax on commercial property value) it had to pay to the town
council. At the times of the respective surveys, these effective
subsidies amounted to approximately 50,000 [pounds sterling] (0.67
[pounds sterling] for each household) per year at Bury and 87,000
[pounds sterling] (1.12 [pounds sterling] for each household) per year
at Luton.
An important decision in any CVM is whether to regard bids (stated
amounts of maximum willingness-to-pay) as relating to the benefits of
the amenity to the individual who responds to the questionnaire or to
the whole household to which he belongs. To draw conclusions from any
CVM sample survey, the mean bid in the sample has to be extrapolated to
the level of the population by scaling up according to either the number
of individuals or the number of households in the population. If
multiplication is according to the number of individuals, the estimate
of aggregate willingness-to-pay will be much higher (twice as high if
there are two persons per household). It is therefore important to
consider whether bids should be interpreted as indications of individual
or household preferences.
Because council tax, which is charged to households rather than
individuals, was our choice of payment vehicle, we framed the question
in terms of how much the respondent was willing to pay on behalf of his
or her household. Subsequently we scaled up the mean bid according to
the number of households in the particular borough. Whether answers by
individuals on behalf of households will represent the preferences of
all the individuals in the household in the same way as if they had each
been questioned individually and individualistically depends on such
factors as respondents' knowledge of other household members'
utility functions, whether respondents are altruistic and how households
make collective consumption choices. Strand (2007) sets out assumptions
that would yield the prediction that it would not matter to the final
aggregate valuation whether individuals were asked to respond as
individuals or as representatives of their households. However, Lindhjem
and Navrud (2009) reported that separate Norwegian samples answering the
same question (about an environmental amenity) made very similar mean
bids whether they were asked to respond in the individual or household
mode. Then it would make a very substantial difference to aggregate
valuation whether the mean response was scaled up according to the
number of individuals or the number of households in the population.
Delaney and O' Toole (2004) asked respondents what would be the
maximum "you" would be willing to pay for provision of Irish
public service broadcasting. A follow-up question then asked how
respondents had interpreted "you" and nearly two-thirds
indicated that they had in fact been thinking of payment on behalf of
the whole household. Scaling up of course would produce a much higher
valuation from a sample where the researchers regarded the response unit
as the individual but, from these earlier findings, there is a danger
that some respondents' bids in fact reflect preferences of all
household members even if the question appeared to be individualistic.
Our decision to ask our respondents for willingness-to-pay on behalf of
their households therefore not only accords with the collective payment
vehicle of council tax but also recognises the need for caution
suggested by the literature. Of course, some respondents asked about
household willingness-to-pay may ignore the preferences of other
household members and our decision to ask about household valuation and
then to scale up by the number of households may therefore introduce a
conservative bias to our valuation of the football clubs.
Measurement
With all these considerations in mind, and after prior cognitive
testing on twenty individuals, our first valuation question in the
surveys was therefore framed as follows:
"Please circle the highest amount you are willing to pay, on
behalf of your household, per month, via an increase in Council Tax, to
guarantee [Bury FC's/ Luton Town's] professional league
status, 0 [pounds sterling] 2 [pounds sterling] 4 [pounds sterling] 6
[pounds sterling] 8 [pounds sterling] 10 [pounds sterling] 12 [pounds
sterling] 14 [pounds sterling] 16 [pounds sterling] 18 [pounds sterling]
20 [pounds sterling]."
As is typical in CVM studies, the majority of answers were expected
to be (and were indeed) zero. The most obvious reason for answering zero
is that the respondent in fact places no value on the amenity. This
would be a 'true zero'. However, Meyerhoff and Liebe (2006)
noted reasons why some zeros might be 'false'. For example,
some respondents might value the amenity positively but give the answer
zero in protest at some aspect of the questionnaire or question. In our
case, the most obvious reason for protest is that using council tax to
pay for football might be considered improper and outside the legitimate
sphere of local government. To attempt to address this potential source
of false zeros, we therefore asked the valuation question again, but
this time including information that the town council already gave
financial support to the football club (implying that additional support
would not create a new precedent). Thus, for Bury residents, the second
valuation question was:
"In reality, Bury Metropolitan Council sponsor Bury FC each
season which equates to 0.67 [pounds sterling] per household per year.
The sponsorship includes the name of Bury Metropolitan Council on the
team shirts, advertisements in the match day programme and tickets
distributed to school children. Considering the extra information,
please circle the highest amount you are now willing to pay on behalf of
your household, per month via an increase in Council tax, to guarantee
Bury FC's professional league status, 0 [pounds sterling] 2 [pounds
sterling] 4 [pounds sterling] 6 [pounds sterling] 8 [pounds sterling] 10
[pounds sterling] 12 [pounds sterling] 14 [pounds sterling] 16 [pounds
sterling] 18 [pounds sterling] 20 [pounds sterling]."
Similarly, the Luton questionnaire also posed a second valuation
question:
"In reality, Luton Borough Council assist Luton Town Football
Club through reduced rates each season equivalent to 1.12 [pounds
sterling] per household per year. Considering the extra information,
please circle the highest amount you are now willing to pay on behalf of
your household, per month via an increase in Council Tax, to ensure in
future Luton Town's professional league status, 0 [pounds sterling]
2 [pounds sterling] 4 [pounds sterling] 6 [pounds sterling] 8 [pounds
sterling] 10 [pounds sterling] 12 [pounds sterling] 14 [pounds sterling]
16 [pounds sterling] 18 [pounds sterling] 20 [pounds sterling]."
The second question, as expected, elicited a greater number of
positive responses: the number of those willing to pay something
increased by exactly one-third in Bury and by 22 per cent in Luton.
Since this second valuation was made in light of fuller information, and
because the nature of the information was likely to have decreased the
incidence of protest zeros, it is responses to this valuation question
that we used when estimating aggregate willingness-to-pay in each
community. Because the question referred to willingness-to-pay through
an increase in council tax, the valuations should be interpreted as
reflecting willingness-to-pay beyond the level of current subsidies. The
current subsidy of 0.67 [pounds sterling] per household in Bury
converted to 49,817 [pounds sterling] at the level of the borough; and
1.12 [pounds sterling] per household in Luton implied a current council
subsidy of 87,136 [pounds sterling].
In addition to the valuation questions, each survey also collected
information on socio-demographic variables, to enable us to form a
judgement on the representativeness of the sample. It also inquired
whether the respondent attended matches at the local football club, to
enable us to distinguish between the willingness-to-pay of users and
non-users.
Data collection
Each of the two surveys was administered by mail, (15) using
samples of 1,200 addresses purchased from a market research company. The
samples were based on the electoral roll for each municipality. (16) The
mailing was directed at one (randomly selected) named adult in each
household. The request to participate was personalised as much as
possible, for example in the greeting (Dear xxxx) and the date at the
head of the letter. Personalisation is likely to raise the response rate
(Dillman, 2007). The covering letter was printed on university headed
notepaper to lend authority and there were clear details on how to
contact a named researcher with any questions. A postage-paid return
envelope was provided. In each of Bury and Luton, 1,200 questionnaires
were sent out. Of these, 27 sent to Bury and 23 sent to Luton were
returned unopened or blank. Exactly half of these fifty were marked to
indicate that the addressee had either died or was no longer resident. A
variety of other reasons, such as poor eyesight preventing completion,
accounted for the other half. With the surveys returned unopened or
blank excluded, potential sample sizes were 1,173 for Bury and 1,178 for
Luton. The number of useable responses was 141 from Bury and 142 from
Luton. Thus response rates were 12.02 per cent and 12.05 per cent
respectively.
Results
This section describes how we used the responses to estimate
aggregate willingness-to-pay for the existence of professional football
in each town. The inputs into the calculations are summarised in tables
2 (Bury) and 3 (Luton).
Fifty-six (39.7 per cent) of the 141 responses from Bury and 76
(53.5 per cent) of the 142 responses from Luton entered a valuation
greater than zero. The greater willingness of residents of Luton to pay
something to aid the town football club over and above the modest level
of current subsidy was evident amongst both users (those who attend any
games) and non-users (those who never attend). Of attendees, 74.2 per
cent in Luton and 58.7 per cent in Bury made a positive bid. Amongst
non-attendees, the proportions were 45.6 per cent and 24.3 per cent
respectively. Mean (monthly) willingness-to-pay conditional on being
willing to pay anything was, however, rather higher in Bury than in
Luton (4.52 [pounds sterling] versus 3.13 [pounds sterling]).
While, as shown in table 1, the samples proved reasonably
representative of the towns' populations in terms of the
demographic variables included in Census data, (17) this does not imply
of course that they were unbiased in respect of the variable we sought
to estimate. Although our covering letter emphasised that responses
would be equally valuable even if the addressee were uninterested in
sport, nevertheless it would be fair to suppose that a disproportionate
number of responses received would be from those with some interest in,
or sympathy with, the subject of the questionnaire, football. Again, the
proportions of responses from those who attended games, even if only
occasionally, seemed quite high relative to mean recorded attendances at
the stadia. It is plausible therefore that the mean value of
willingness-to-pay in the samples (once the zero bids are included, the
means for monthly willingness-to-pay were 1.79 [pounds sterling] in Bury
and 1.68 [pounds sterling] in Luton) is an upwardly biased estimate of
the mean willingness-to-pay in the whole population. Extrapolating to
the aggregate valuation of the population by multiplying the sample mean
by the number of households would therefore be likely to yield an
exaggerated estimate of community benefit from the football club. It
would be very much an upper-bound estimate of the value of the amenity
because it would be based on the assumption that, on average,
non-respondents valued the football club equally with respondents.
Correcting for this problem is difficult because no information is
available on non-respondents. Johnson and Whitehead (2012) note that a
common solution in the CVM literature is to attribute a valuation of
zero to all those who did not respond to the survey and then to treat
the resulting calculation of population aggregate willingness-to-pay as
a lower-bound estimate, to be set alongside the upper-bound estimate. We
also follow this procedure. It has the additional benefit of obviating
the need to be concerned over any under-representation of particular
demographic groups with low interest in football. Since the initial
sampling is random, and if most members of demographic groups fail to
respond because they are indifferent to football, their views still
count if a valuation of zero is imputed to them.
A disadvantage of attributing zero willingness-to-pay to
non-respondents is that, given that response rates tend to be low,
particularly for mailed surveys, it typically yields a lower-bound
estimate that is very far from the corresponding upper-bound estimate.
Of course, obtaining a wide range for the estimates is a function of
adopting two extreme and opposite assumptions, namely that
non-respondents do not value the amenity at all and that non-respondents
value the amenity exactly as much as respondents.
To calculate the lower-bound estimate of aggregate town
willingness-to-pay for Bury, we began with the mean willingness-to-pay
of 1.79 [pounds sterling] observed for the 141 respondents. We imputed a
mean of zero to the 1,032 non-respondents. Across the whole potential
sample of 1,173 persons, this gave an overall mean monthly
willingness-to-pay of 0.22 [pounds sterling].
Adopting the conservative assumption that each individual response
represents the valuation of the respondent's whole household, we
multiplied the mean of 0.22 [pounds sterling] by 74,335, which was the
number of households in Bury, according to Office of National Statistics
estimates. This yielded an estimate of aggregate willingness-to-pay of
16,033 [pounds sterling] per month across households resident in
Bury--or 192,396 [pounds sterling] per year. Adding in the implicit
subsidy already paid by the municipality (recall that the responses
concerned willingness-to-pay beyond the level of the current subsidy)
yields our lower-bound estimate for community benefit (from Bury having
a professional football team) of 242,200 [pounds sterling].
Table 2 sets out the details of the calculations and also those for
the upper-bound estimate. The upper-bound estimate is 1,650,298 [pounds
sterling] per year.
Table 3 sets out details of the equivalent calculations for Luton.
Here the lower- and upper-bound estimates are 275,758 [pounds sterling]
and 1,651,902 [pounds sterling] respectively. Thus, estimates prove to
be of very similar orders of magnitude in the two cases.
Although we have followed precedent in the literature by presenting
lower- and upper-bound estimates, our policy discussion will rely only
on the lower-bound estimates. This is consistent with other, earlier
decisions, noted above, that deliberately adopted conservative
assumptions in the procedures we followed. If estimates are used to
assess the case for public intervention in favour of the survival of
small town football clubs, respect for taxpayers requires that the case
not be exaggerated. Choosing lower-bound estimates eliminates any
contamination due to differential response rates across groups defined
by, say, demography or attendance at the football games. Thus, males,
for example, are over-represented in the Bury sample. But, when we
attribute a value of zero to willingness-to-pay to all non-respondents,
this restores the missing females to the data set on which the
lower-bound estimate is based. It is therefore truly the lower-bound
estimate.
The annual community benefits of 242,200 [pounds sterling] and
275,758 [pounds sterling] at Bury and Luton respectively represent value
produced by the football club which is not captured by the club in its
revenue from ticket sales. Instead it accrues as either consumer surplus
(for attendees) or to attendees or non-attendees in the form of
externalities (providing a focus for discussion with neighbours, civic
pride, and so on).
To be sure, amounts of this order of magnitude would represent
small change in the worlds of the English Premier League and other
top-tier European competitions. For example, Castellanos et al. (2011),
in the only earlier study on football similar to ours, found an annual
benefit of approximately 3m [euro] from the presence of a First Division
club in a Spanish city. This is many times our estimates. But their
figure was nevertheless lower than 5 per cent of the budget of the
football club.
By contrast, our more modest estimates of benefit to the small
towns of Bury and Luton represent amounts that are somewhat higher
relative to the revenues of the respective clubs. Unfortunately the
exact revenues of the two clubs for the respective seasons were
unavailable to us. Each club is registered with a form of ownership
where accounts have not needed to be published since 2001 and each
declined to supply us with their private data. However, the accounting
firm Deloitte estimated that the 'average' club in the
Division had revenue of 2.7 million [pounds sterling] in 2008 and 2.8
million [pounds sterling] in 2009. Our estimates of community benefits
not captured by the clubs represented 9 per cent (Bury) and 10 per cent
(Luton) of these average figures (18) (and a much higher proportion of
average ticket revenue). (19) Moreover, the community benefits appear
large in the context of average losses in the division of 300,000
[pounds sterling] and 400,000 in the two years (Deloitte, 2010). If the
results from these towns can be generalised (and the similarity of the
two estimates encourages us to believe that they may be capable of being
generalised), then it is clear that it would be plausible for situations
to arise where it would be socially efficient for an intervention to be
made to ensure that a small town professional club survived a threat to
its ongoing operation. Our results suggest that extending professional
football to a fourth tier, to serve smaller centres, does not produce a
structure that is 'too large'. Returning to the particular
cases of Bury and Luton, the subsidies paid to their football club by
the respective town councils were in each case certainly significantly
below the value the club created for residents.
Finally in this section, it is of interest to compare the relative
sizes of bid made by attendees and non-attendees in the Spanish sample
as reported by Castellanos et al. (2011) and in our samples from Bury
and Luton. In the Spanish case, the mean bids were quite similar, 10.97
[euro] and 9.30 [euro]. But in Bury/Luton, the mean for attendees (2.87
[pounds sterling]/3.28 [pounds sterling]) was much higher than that for
non-attendees (0.92 [pounds sterling]/1.07 [pounds sterling]). Although
there are caveats in comparing results across the studies (attendees
represented a much higher proportion of the sample in Spain; response
rate was much higher because the researchers' budget allowed
face-to-face interviewing; and face-to-face interviewing itself tends to
raise mean willingness-to-pay, according to Noonan, 2003), this is
nevertheless suggestive of the possibility that externalities from
football fall off faster than the benefit from active consumption as one
considers lower levels of football. But even so, we should reiterate,
community benefit is still high relative to the size of business of the
football club.
Conclusion
Our contingent valuation study has followed precedents in the
academic literature in seeking to evaluate the benefit to a community
from hosting a professional sports team. It is worth reiterating that we
have been consistently conservative when making methodological choices
in the sense of always choosing the option that would tend to yield a
lower estimate of aggregate benefit. First, and this was a choice
imposed by budget constraints, we used a postal rather than a
face-to-face survey. Second, we attributed a zero valuation to every
single non-respondent. Third, we assumed that the 'bid' of
each respondent reflected the club's value to his or her whole
household rather than to him or her individually. The estimates of
aggregate value are therefore clearly lower-bound estimates.
Previous studies have not always been so conservative, in
particular some previous work (such as Castellanos et al., 2011) perhaps
risks exaggerating benefits by extrapolating results from questioning
individuals by scaling up by the number of individuals rather than by
the number of households in the city. Nevertheless, our results can
still only reasonably be interpreted as showing rather lower
willingness-to-pay compared with earlier case studies; for example our
(lower-bound) estimates of mean willingness-to-pay are under 2 [pounds
sterling], compared with about 10 [euro] in the case of a study of a
Spanish club, Castellanos et al, 2011. But all previous studies, in
whatever sport, related to major league clubs whereas the two clubs in
our case studies operate at the lowest level of professional football.
What is significant for sports policy is that benefits generated for the
communities prove to be higher than in previous studies when evaluated
relative to the economic size of their club. If these case studies are
representative of what would be found across bottom-tier professional
football, it is evident that there is a case on standard welfare
economic grounds for municipalities to consider intervention in the
event that the survival of their town club comes into question. In times
of fiscal pressure, intervention might necessarily take other forms than
direct subsidy, for example protection through planning restrictions on
the use to which a developer bidding for the land on which the stadium
sits could put it, or promotion of a new mode of ownership. Beyond the
implications for individual communities, a wider point suggested by our
analysis is that any attempt in the future to reduce the size of the
professional football league structure in England would not be without
significant social cost.
REFERENCES
Alexander, D.L., Kern, W. and Neill, J. (2000), 'Valuing the
consumption benefits from professional sports franchises', Journal
of Urban Economics, 48, pp. 321-37.
Atkinson, G., Mourato, S., Szymanski, S. and Ozdemiroglu, E.
(2008), 'Are we willing to pay enough to 'back the bid'?
Valuing the intangible impacts of London's bid to host the 2012
Summer Olympic Games', Urban Studies, 45, pp. 419-44.
Baade, R.A. (2006), 'A primer on the economic impact of
mega-sports events and investments in sports infrastructure', in
Rodriguez, P., Kesenne, S. and Garcia, J. (eds), Sports Economics After
Fifty Years: Essays in Honour of Simon Rottenberg, pp. 51-68, Oviedo,
Ediciones de la Universidad de Oviedo.
Buraimo, B., Forrest, D. and Simmons, R. (2007), 'Freedom of
entry, market size and competitive outcome: evidence from English League
soccer', Southern Economic Journal, 73, pp. 204-13.
Castellanos, P., Garcia, J. and Sanchez, J.M. (2011), 'The
willingness to pay to keep a football club in a city: how important are
the methodological issues?', Journal of Sports Economics, 12. pp.
464-86.
Ciriacy-Wantrup, S.V. (1947), 'Capital returns from soil
conservation practices', Journal of Farm Economics, 29, pp.
1181-96.
Delaney, L. and O'Toole, F. (2004), 'Irish public service
broadcasting: a contingent valuation analysis', Economic and Social
Review, 35, pp. 321-50.
Deloitte (2010), Annual Review of Football Finance, 19th Edition,
Manchester, Deloitte.
Dillman, D.A. (2007), Mail and Internet Surveys: The Tailored
Design Method, 2nd edition, Hoboken, NJ, John Wiley.
Fenn, A.J. and Crooker, J.R. (2009), 'Estimating local welfare
generated by a professional sports team: an application to the Minnesota
Vikings under threat of relocation', Southern Economic Journal, 76,
pp. 198-223.
Forrest, D., Simmons, R. and Feehan, P. (2002), 'A spatial
cross-sectional analysis of the elasticity of demand for soccer',
Scottish Journal of Political Economy, 49, pp. 336-55.
Hakes, J., Turner, C.S. and Hutmaker, K. (2011), 'I don't
care if I never get back? Time, travel costs, and the estimation of
baseball season-ticket demand', International Journal of Sport
Finance, 6, pp. 119-37.
Johnson, B.K., Groothuis, P.A. and Whitehead, J.C. (2001),
'Value of public goods generated by a major league sports team: the
CVM approach'. Journal of Sports Economics, 2, pp. 6-21.
Johnson, B.K., Mondello, M.J. and Whitehead, J.C. (2007), 'The
value of public goods generated by a National Football League
team', Journal of Sport Management, 21, pp. 123-36.
Johnson, B.K. and Whitehead, J.C. (2012), 'Contingent
valuation of sports', in Shmanske, S. and Kahane, L.H. (eds), The
Oxford Handbook of Sports Economics, vol. 2, pp. 335-49, Oxford, Oxford
University Press.
Lindhjem, H. and Navrud, S. (2009), 'Asking for household or
individual willingness to pay for environmental goods? Implications for
aggregate welfare measures', Environmental and Resource Economics,
43, pp. 11-29.
Loomis, J. (2011), 'What's to know about hypothetical
bias in stated preference valuation studies?', Journal of Economic
Surveys, 25, pp. 363-70.
Meyerhoff, J. and Liebe, L. (2006), 'Protest beliefs in
contingent valuation: explaining their motivation', Ecological
Economics, 57, pp. 583-94.
Murphy, J.J., Allen, P.G., Stevens, T.H. and Weatherhead, D.
(2005), 'A meta-analysis of hypothetical bias in stated preference
valuation', Environmental and Resource Economics, 30, pp. 313-25.
Noll, R.G. and Zimbalist, A. (eds) (1997), Sports, Jobs and Taxes:
The Economic Impact of Sports Teams and Stadiums, Washington DC,
Brookings Institution.
Noonan, D.S. (2003), 'Contingent valuation and cultural
resources: a meta-analytical review of the literature', Journal of
Cultural Economics, 27, pp. 159-76.
Preuss, H. and Werkman, K. (2011), 'Erlebniswert Olympischer
Winterspiele in Munchen 2018', Sport und Gesselschaft, 8, pp.
97-123.
Strand, J. (2007), 'Public good valuation and intra-family
allocation', Environmental and Resource Economics, 38, pp. 527-43.
Szymanski, S. (2014), 'Insolvency in English football',
in Goddard, J. and Sloane, P. (eds), Handbook on the Economics of
Professional Football, pp. 100-16, Cheltenham, Edward Elgar.
Tabner, B. (1992), Through the Turnstiles, Harefield, Middlesex,
Yore Publications.
Walker, M. and Mondello, M.J. (2007), 'Moving beyond economic
impact: a closer look at the contingent valuation method',
International Journal of Sport Finance, 2, pp. 149-60.
Walton, H., Longo, L. and Dawson, P. (2008), 'A contingent
valuation of the 2012 London Olympic Games 'Journal of Sports
Economics, 9, pp. 304-17.
Wicker, P., Hallmann, K., Breur, C. and Feiler, S. (2012a),
'The value of Olympic success and the intangible effects of sports
events --a contingent valuation approach in Germany', European
Sport Management Quarterly, 12, pp. 337-55.
Wicker, P., Prinz, J. and von Hanau, T. (2012b), 'Estimating
the value
of national sporting success', Sport Management Review, 15, pp.
200-10.
NOTES
(1) We reviewed the potted histories of all clubs ever to have
played in the Football League, included in Tabner (1992). Several clubs
present in 1924-5 failed but only Merthyr Town, New Brighton and South
Shields were not reformed to continue to the present day (there is a
club named Merthyr Town playing today but it was not founded until 2010
whereas the former Football League club was dissolved in 1934; the case
is one of enthusiasts adopting a historic name rather than survival from
adverse circumstances). However, some clubs, such as Bradford Park
Avenue, liquidated in 1974, were reformed only to play in much lower
level competition.
(2) Maidstone United was reformed after a lag and today plays in
the seventh tier of English football (the structure is explained
atwww.fansfocus.com, accessed December 15, 2014).
(3) Barrow was voted out of the Football League in 1972 and
currently plays in the sixth tier of the structure; Southport finished
second-bottom in the Football League three years running before ejection
in 1978. It currently plays, with part-time professional personnel, in
the fifth tier (histories from the clubs' respective Wikipedia
entries).
(4) Currently the two worst performing clubs each season have to
swap places with the top two performers in the semiprofessional feeder
league.
(5) Strictly, under the compensated demand curve.
(6) In microeconomics texts, this species of market failure is
typically illustrated by reference to public utilities. As with
performance goods, public utilities typically feature fixed costs that
are high relative to the maximum revenue that can be extracted from the
market.
(7) Support for the relevance of external benefits from a
professional club was provided, for instance, in Johnson et al. (2001).
In a sample of residents in Pittsburgh, 44 per cent reported reading
about, and 34 per cent talking about, the local ice hockey team at least
a few times per week in-season. Participation in such activity was
positively correlated with willingness-to-pay to ensure continuation of
the sport in the city.
(8) Willingness-to-pay is a measure of the equivalent variation. It
measures the loss of income which, to the respondent, would be
equivalent, in terms of decrease in utility, to loss of the amenity (the
football club).
(9) Note that users and non-users alike consume the public good
because, for example, both groups enjoy talking about the last result.
Of course, the typical valuation of this aspect of the club's
output may differ as between the two groups.
(10) An example of strategic behaviour would be if a respondent
reasoned that he would benefit from an amenity but that he would not
bear any of the cost because he was not a taxpayer. His incentive is to
overstate willingness-to-pay because the results of the survey might be
used to determine that the amenity will be provided, to his advantage.
(11) Luton Town in fact lost its relegation fight and its 89 years
tenure in the professional league. The season was quite successful in
terms of playing record but the efforts of the team in the end failed to
overcome the deduction of 30 league points that had been imposed earlier
for financial irregularities and violation of player transfer rules.
(12) In the respective seasons during which the surveys took place,
mean attendance at Bury home matches was 2,594 and that at Luton 5,968
(www.tonykempster.co.uk).
(13) Constraining answers to be zero or positive implies an
assumption that no residents would prefer the club to lose its place in
the league. We speculate that this is more plausible than for clubs in
major leagues. Generally, fourth-tier football does not attract
attendances of a size that cause congestion and social disruption.
(14) A disadvantage is that there may be many residents with a low
valuation, here for example below 1 [pounds sterling]. Given that they
are likely to enter zero, this will be a source of downward bias in our
estimate of aggregate willingness-to-pay.
(15) Noonan (2003), on the basis of a meta-analysis of CVM studies
in the field of culture (including sport), reports that mail surveys
tend to yield lower willingness-to-pay estimates than alternatives,
especially doorstep interviews. Cost dictated our choice to sample
households by mail questionnaire. Noonan's finding implies the
possibility that this will yield a conservative valuation.
(16) Valuation of the amenity provided by the football club will be
underestimated to the extent that it will not include benefits to
persons residing beyond the local authority boundary (who may
nevertheless be close enough to the town to feel an affinity with it and
its football club).
(17) One exception, in Bury only, was the preponderance of male
responses, nearly 74 per cent (the proportion was 50 per cent in Luton).
Also, in both towns, retirees were over-represented and those
self-identifying as other than 'white British' were
underrepresented. Part of this latter discrepancy may be associated with
a low propensity for members of minority communities to register on the
electoral roll.
(18) It is likely that Luton in fact had greater revenue, and Bury
lower revenue, than the 'average' club in the Division, based
on average attendances compared with other clubs (www.tonykempster.
co.uk).
(19) At an 'average' club, ticket sales including season
tickets were said to represent about 40 per cent of total revenue, the
remainder coming from sponsorship, merchandising and television payments
made to the League.
Andy Barlow and David Forrest, University of Liverpool Management
School. E-mail:
[email protected]
Table 1. Comparison of borough and sample
characteristics
Bury
Borough Sample
Population 183,300 --
Area 99.48 km2 --
Density 1 843 / km2 --
Households 74,335 --
Male 0.487 0.736
Female 0.513 0.264
Full time 0.429 0.447
Part time 0.120 0.071
Self-employed 0.081 0.092
Unemployed/inactive 0.179 0.004
Student 0.058 0.014
Retired 0.133 0.341
O level/GCSE 0.321 0.248
A Level 0.200 0.213
Degree 0.180 0.277
No qualification 0.292 0.262
White 0.921 0.986
Mixed race 0.000 0.000
Asian or Asian British 0.054 0.014
Black or black British 0.016 0.000
Chinese or other ethnic group 0.008 0.000
Luton
Borough Sample
Population 188,800 --
Area 43.35km2 --
Density 4355 / km2 --
Households 77,800 --
Male 0.507 0.503
Female 0.493 0.497
Full time 0.425 0.394
Part time 0.104 0.098
Self-employed 0.067 0.077
Unemployed/inactive 0.203 0.114
Student 0.093 0.042
Retired 0.109 0.275
O level/GCSE 0.288 0.299
A Level 0.170 0.218
Degree 0.150 0.183
No qualification 0.313 0.300
White 0.680 0.845
Mixed race 0.028 0.014
Asian or Asian British 0.193 0.064
Black or black British 0.079 0.007
Chinese or other ethnic group 0.020 0.007
Source: Borough population statistics
www.neighbourhood.statistics.gov.uk.
Table 2. Summary statistics and calculation of community benefit: Bury
Sample size 141
Potential sample size 1,173
Number of responses with wtp>0 56
Number of responses with wtp=0 85
Mean wtp where wtp>0 4.52 [pounds sterling]
Mean wtp across all responses 1.79 [pounds sterling]
Mean monthly wtp of potential sample Lower-bound
(assume non-responses wtp=0)
Mean monthly wtp of potential sample 0.22 [pounds sterling]
(assume non-responses wtp=1.79
[pounds sterling])
(Multiply by number of households in 16,033 [pounds sterling]
Bury, 74,335)
Estimated wtp across all households in
Bury (Multiply by 12 to convert to 192,336 [pounds sterling]
annual wtp)
Estimated annual wtp across all
households in Bury
(Add) value of current annual subsidy 49,804 [pounds sterling]
evaluated at 0.67 [pounds sterling]
per household
Estimate of annual benefit to Bury 242,200 [pounds sterling]
households
Sample size
Potential sample size
Number of responses with wtp>0
Number of responses with wtp=0
Mean wtp where wtp>0
Mean wtp across all responses
Mean monthly wtp of potential sample Upper-boun
(assume non-responses wtp=0)
Mean monthly wtp of potential sample 1.79 [pounds sterling]
(assume non-responses wtp=1.79
[pounds sterling])
(Multiply by number of households in 133,375 [pounds sterling]
Bury, 74,335)
Estimated wtp across all households in
Bury (Multiply by 12 to convert to 1,600,494 [pounds sterling]
annual wtp)
Estimated annual wtp across all
households in Bury
(Add) value of current annual subsidy 49,804 [pounds sterling]
evaluated at 0.67 [pounds sterling]
per household
Estimate of annual benefit to Bury 1,650,298 [pounds sterling]
households
Notes: wtp signifies willingness-to-pay (monthly unless stated
otherwise). Individuals' wtp is shown to the nearest penny but 'exact'
means were used in subsequent calculations.
Table 3. Summary statistics and calculation of community benefit:
Luton
Sample size 142
Potential sample size 1,178
Number of responses with wtp>0 76
Number of responses with wtp=0 66
Mean wtp where wtp>0 3.13 [pounds sterling]
Mean wtp across all responses 1.68 [pounds sterling]
Lower-bound
Mean monthly wtp of potential sample 0.20 [pounds sterling]
(assume non-responses wtp=0)
Mean monthly wtp of potential sample
(assume non-responses wtp=1.79
[pounds sterling])
(Multiply by number of households in 15,719 [pounds sterling]
Luton, 74,335)
Estimated wtp across all households in
Luton
(Multiply by 12 to convert to annual 188,622 [pounds sterling]
wtp)
Estimated annual wtp across all
households in Luton
(Add) value of current annual subsidy 87,136 [pounds sterling]
evaluated at 0.67 [pounds sterling]
per household
Estimate of annual benefit to Luton 275,758 [pounds sterling]
households
Sample size
Potential sample size
Number of responses with wtp>0
Number of responses with wtp=0
Mean wtp where wtp>0
Mean wtp across all responses
Upper-bound
Mean monthly wtp of potential sample
(assume non-responses wtp=0)
Mean monthly wtp of potential sample 1.68 [pounds sterling]
(assume non-responses wtp=1.79
[pounds sterling])
(Multiply by number of households in 130,397 [pounds sterling]
Luton, 74,335)
Estimated wtp across all households in
Luton
(Multiply by 12 to convert to annual 1,564,766 [pounds sterling]
wtp)
Estimated annual wtp across all
households in Luton
(Add) value of current annual subsidy 87,136 [pounds sterling]
evaluated at 0.67 [pounds sterling]
per household
Estimate of annual benefit to Luton 1,651,902 [pounds sterling]
households
Notes: wtp signifies willingness-to-pay (monthly unless stated
otherwise). Individuals' wtp is shown to the nearest penny but 'exact'
means were used in subsequent calculations.
