Ernst Jonsson Does It Pay to Keep Shops Open in Sparsely Populated Areas? A Cost-Benefit Analysis of a Hypothetical Shop Closure ABSTRACT. If the only shop in a rural locality is closed down the households will be faced with increased purchasing costs and other welfare losses. On the other hand the closure of the shop would release labour and other resources (e.g., premises, fittings, and stocks). Consequently the economic result is equal to the difference between those costs and revenues (the alternative value of these resources) that do not arise if the shop remains open. Whether society as a whole gains or loses from a shop closure will depend in the first place on (a) how far the households have to travel to the nearest alternative shop; (b) how many households are effected by the closure; and (c) what value the resources thus released could have in an alternative use. Those rural shops that received government investment support in Sweden are located on average 15 km from the nearest competitor and serve 100 households. The typical shop is also located in an area where there is a permanent shortage o f jobs. According to the analysis of this paper, a proposed closure o f such a shop would involve a social loss o f approximately 130,000-140,000 SKR per year.
THE PROBLEM OF SHOP SUPPORT
In a locality with a small or falling number of customers it can be difficult for the shop-owner to earn a normal return on the labour and capital he has invested. For this reason the shop in such a locality is in danger of being closed down, with the result that households would be obliged to buy foodstuffs and other everyday necessities in some other way. Households would be obliged to adopt purchasing methods that are more expensive (for example, car journeys to another locality) or are valued less highly (for example, home deliveries) than shopping at their own local shop. If the only shop in the locality is closed down households would therefore suffer losses. If the shop-owner is to be provided with a true economic basis for 'assessing his situation, the benefits (in the form of losses avoided) accruing to households as a result of continued operations have to be included in his financial calculation. In the perfect market economy this question would be settled by means of bargaining between the shop-owner and the households in the locality. As far as the households are concerned it would be rational, for example, for them to offer the shop-owner a fixed, annual remuneration (per household) if he were to take on the responsibility of keeping the shop open. The welfare losses the households avoid would then Journal of Consumer Policy 6 (1983) 4 3 7 - 4 5 6 . 0 3 4 2 - 5 8 4 3 / 8 3 / 0 0 6 4 - 0 4 3 7 $02.00. © 1983 by D. Reidel Publishing Company.
438
Ernst Jonsson
determine the maximum amount they were willing to pay. On his side the shop-owner would require the households to compensate him for his labour and risk-taking in keeping his shop open. Thus, in theory the question is settled by means of free bargaining between the shop-owner and the households affected. In practice, however, it is doubtful whether such a solution via bargaining is feasible because: neither the shop-owner nor the households are aware of such a possibility, or they consider it impracticable; - or the households feel it is up to the local authority or the state to guarantee the continued existence of the shop. Where the continued existence of a rural shop cannot be guaranteed by means of a market solution there are economic arguments in favour of the state providing the shop with a regular subsidy. Such a subsidy may be regarded as remuneration paid by the state to the shop-owner - as a representative of the households - in return for his keeping the shop open. This means that what is economically profitable will also be profitable from a private business point of view. However, to achieve this, the subsidy must not be less than what is needed to make the shop-owner willing to continue, though on the other hand the subsidy must not be larger than the difference between the benefit to society and to the shop. -
THE GOAL OF THE STUDY
Undoubtedly society can profit if shops in sparsely populated areas are not closed down, and this is why the Swedish government for some years now has been providing financial support for the rebuilding or extension of such shops. How much does society really gain from ensuring that shops in sparsely populated areas remain in business? This study seeks to answer the question by taking as the starting point a hypothetical closure of a rural shop which is typical in important respects and which has been receiving government investment support.
CHOOSING A TYPICAL SHOP IN A SPARSELY POPULATED AREA
Those rural shops that received government investment support during the period 1/7 1973 to 30/6 1978 were located on an average 15 km from the nearest competitor and 35 km from the nearest urban centre. The shop at Vormsele in northern Sweden is typical in these respects, though there was yet another reason why this shop
Costs and Benefits of a Shop Closure
439
was selected as the object of the study. It was in fact closed from December 1976 to March 1978. Thus, when this enquiry was carried out in 1979, households had had recent experience of the closure of a shop. These quasi-experimental conditions make the Vormsele case well suited for this study.
METHODOLOGY AND DATA USED
Cost-Benefit Analysis The purpose of a cost-benefit analysis is to show whether society as a whole achieves greater welfare as a result of launching a particular project rather than abstaining from it. It is assumed that society's welfare increases if the social benefits of the project exceed the social costs. What then are social costs and benefits? Social costs represent the value of the resources in the best alternative use, i.e., the opportunity cost, for there will be a sacrifice involved in applying resources to one use rather than another. Normally the opportunity cost is calculated on the basis of the prevailing market prices of the various resources (J6nsson, 1976; pearce, 1971). According to welfare theory the social benefit is equal to people's willingness to pay for the project. Willingness to pay may be estimated directly (via, for example, interviews) or indirectly (via observations of the way the individual actually behaves). In both cases the aim is to arrive at a kind of demand curve for the service that is being evaluated (Bohm, 1972).
The Indirect Method of Estimation The indirect method of estimating willingness to pay is used in costbenefit analysis of road or other transport projects. As an example of such a project Mishan (1972) mentions a tunnel under the English Channel. Such a tunnel would shorten travel times compared with those entailed in transport by ferry or boat. The transport cost er journey would thereby fall from C1 to C~, which would lead toPaen increase in the number of cross-channel travellers from OM 1 to OM 2 . The benefits of the tunnel may be seen in Figure 1: A fall in the unit cost from OC] to OC2, when the journey is undertaken by tunnel, implies a cost-saving of C2 C1 for each of the OM1 journeys now diverted from the existing means of travel to the tunnel. The area of rectangle D represents, therefore, a gain from the diverted traffic. At a lower unit cost of the journey, OC2, MIM2 journeys are -
-
440
Ernst Jonsson
undertaken that would not otherwise take place. These additional journeys are the generated traffic, and the gain they offer is measured by the consumer surplus triangle G. The sum of the rectangle D and the triangle G is an operational measure of the maximum amount consumers are willing to pay in order to be able to use the tunnel.
M1
M2
Fig. 1. The benefits of a tunnel project.
In order to assess the consumer's willingness to pay for the continued existence of the local shop use is first made of a method of estimation similar to that used for the road project. In that case, an estimate of the welfare loss suffered by households in the event of a hypothetical shop closure may be obtained by noting how the number of purchases changes when the cost of each purchase increases in terms of distance travelled and time (Computational Model A). However, this method does not take account of the fact that the composition of purchasing methods (for example, shop visits, home deliveries) changes if the local shop is closed down. A more reliable method is therefore developed based on a step-by-step choice of purchasing methods (Computational Model B).
Purchasing Habits of Households Before and Following Closure of the Local Shop The shop at Vormsele serves 55 households comprising 115 persons. Each one of these households was interviewed regarding its purchasing habits. At the present time 85% of foodstuffs and other everyday commodities are bought at the local shop. However, if the shop were to be closed, none of the households would move from the locality. They would instead reduce their purchases from three to two times per week, most of these from the shop located some 15 km distant, on a home delivery basis in the main. The purchasing habits would on the whole revert to that which prevailed when the local shop was in fact closed. Consequently, what households experienced when they had no local shop is reflected in what they imagine they would do if the shop were once again to be closed (cf. Table I).
441
Costs and Benefits of a Shop Closure
TABLE I Average N u m b e r of Purchases per Household per Week
Present situation Visit to shop nearest shop while travelling to and from work at urban centre
Probable situation without a shop
2.80
Actual situation without a shop
1.05
1.20
2.40
0.30
0.30
0.15
0.45
0.50
0.25
0.30
0.40
Home deliveries
0.15
0.80
0.80
Total
2.95
1.85
2.00
The majority of the households interviewed prefer to be able to shop locally rather than at a shop located elsewhere and on average each household gave two or three reasons for this preference. The most common reason is the saving in time and travel costs that a local shop enables them to make, and virtually every household made this point directly or indirectly. The majority of households also pointed to the social importance of a local shop (as a meeting place, etc.). Similarly, 40% of the households emphasized the advantage of being able to see what they are buying - in contrast to the situation when they have to rely on home deliveries. Cost of Travel and Time per Purchase
When making purchases the travel cost of the household consists of the cost of using a private car (running costs, taxation, insurance, depreciation), which varies according to the distance travelled. The price the household pays for being able to make its purchases also includes time-related costs: - time taken travelling to and from the shop - the time actually spent making the purchase. This time could be estimated at 20 minutes per visit to the shop (not including the time devoted to social contacts) and 10 minutes on home deliveries. When estimating the time costs it is assumed that the consumer's time is scarce, and this means that the consumer is prepared to make some sacrifice in order to cut down, for example, his travelling time. To this end the consumer might prefer a faster, more expensive mode of travel to a slower and cheaper one. Thus, it can be argued that the consumer implicitly estimates the saving in time at the
442
Ernst Jonsson
difference in cost between the different modes of travel. Thus the consumer's valuation of time is revealed by actual behaviour (Harrison & Quarmby, 1972). In accordance with what has been observed during investigations of leisure travel the non-working time of the consumer is usually valued at between 15 and 20% of the hourly wage, which corresponds to 5 SKR per hour. The cost to the household of different purchasing methods is shown in Table II. This cost, in conjunction with the number of purchases before and after the closure, serves as a stating point for estimating the household's willingness to pay for being able to make purchases of everyday goods. This willingness to pay minus the social cost of the various purchasing methods represents the net benefit households receive from the purchases made. TABLE II Private and Social Cost of Various Modes of Purchasing (SKR)
Cost per visit to nearest shop Cost per shop visit while travelling to and from work Cost per shop visit at urban centre Cost per home delivery Cost per average mode of purchasing of which cost of travel
The present situation with a local shop
The hypothetical situation without a local shop
Private
Social
Private
Social
cost
cost
cost
cost
3.4
3.5
20.4
22.5
1.4
1.4
1.4
1.4
60.9 0.9
69.4 2.8
60.9 0.9
69.4 4.0
8.1 5.4
9.1 6.2
14.0 11.2
18.0 14.3
As will be seen from the table the cost per purchasing method is greater from the point of view of society than from that of the household, which is connected with: - the fact that home deliveries cost the shop more compared with visits to the shop; - the fact that the cost of transport is greater for society (= 7.20 SKR per 10 kin) than for the individual (= 6.20 SKR per 10 kin). The fact is that the car tax and insurance premiums paid by households owning cars do not fully cover the social costs of traffic accidents (that is to say, costs of medical treatment and loss of production).
Costs and Benefits of a Shop Closure
443
Estimation Based on Quasi-Experimental Experiences The cost of travel and time spent is regarded as the price the household pays for being able to complete its purchase. If the local shop is closed down this price will increase, but if the shop is reopened it will instead fall. By reacting to these price changes the households indicate their willingness to pay for access to a local shop. It is therefore possible to deduce this willingness to pay by studying the purchasing behaviour of households prior to, during, and following the actual closing down of the shop at Vormsele. With the closing down of the local shop in December 1976 the fifty or so households no longer had a choice between shopping at a local shop and a shop located elsewhere. Instead they were limited to choosing between: - making their purchases on their way to or from work or undertaking separate shopping visits shopping in a neighbouring locality or in the urban centre - visiting the shop personally or arranging for home deliveries. In this situation with - as a result of the closure - restricted choice the households increased their purchases via home delivery or while travelling to or from work. At the same time there was a fall in the number of visits to shops and in the number of purchases made. The new pattern meant: - purchasing everyday commodities twice a week; purchasing via home delivery 0.8 times a week; - purchasing while travelling to and from work 0.5 times a week; purchasing in the urban centre 0.4 times a week; purchasing at the nearest shop (that is, at the now nearest shop) 0.3 times a week. By this time the households had experience of both having a local shop and being without one. In other words, it may be assumed that they were relatively well informed about the advantages and disadvantages of the various purchasing alternatives. When the shop reopened in March 1978, the households chose to revert to their purchasing pattern before the closure of the shop. Having once again a wider choice the households decided to shop three times a week on average, mainly by paying personal visits to the local shop. Thus, through their own behaviour the households had confirmed that they preferred to be able to do their own shopping every other day at a shop in their immediate locality, even if doing so would give them additional costs in time and travelling compared with an alternative involving fewer visits to a shop (as during the period of closure). This state of aiCfairs may also be taken as a starting point for estimating the loss of welfare that households would suffer if the shop was again -
-
-
-
444
Ernst Jonsson
closed down. In such a situation without a local shop the households would in fact revert to the purchasing pattern that prevailed while the shop was closed. Provided that the households weigh the various alternatives against one another, this pattern of behaviour indicates how much they value a local shop.
RESULTS
Computational Model A: Number of Purchases Before and Following the Supposed Closure If the local shop is closed down there is an increase from 8 to 14 SKR in the cost per shopping visit borne by the households in terms of time and travel (see Table II for specifics). It leads to the number of purchases per household per week falling from 2.95 to 1.85 (Figure 2). There will be 1.1 fewer visits per week. The willingness of households to devote time and travel costs to visiting the shop may be taken as a measure of the value they place on them. After the supposed shop closure this "willingness to pay" is less than the cost of travel and time (= 14 SKR per visit) with which the household is faced. As long as the local shop is there, the willingness to pay for the 1.1 visits is, on the other hand, greater than the corresponding cost per visit (8.1 SKR). Compared with the present situation the welfare of the household therefore falls by the area BCD (= the willingness to pay minus the cost of the shopping) if the local shop is closed down. Since the cost of purchase is greater from the point of society than from that of the household, the social loss is somewhat less and is equal to the dotted area minus the area shaded black: (Eq. 1)
1.1 (
8.1+14.0 2 willingness to pay for visit
9.1 cost of travel and time per visit
) X 52
~ 110 SKR per household per year.
number of weeks
As regards the purchases which continue to be made after the closure the social cost per purchase also increases, from 9.1 to 18 SKR. This corresponds to a loss per household per year of (= shaded area in Figure 2): (Eq. 2)
1.85 (18.0 - 9.1) × 52 ~ 860 SKR
445
Costs and Benefits o f a Shop Closure
Cost per purchase
14.0
C
i t J t I
1.85 After
2.95 Numberer of purchases Before per household per week
Fig. 2. Cost in travel and time per average mode of purchasing, and n u m b e r o f purchases per household per week, before and after closure of the local shop. In SKR.
To sum up, the deterioration suffered by the average household through the closure of the shop is converted into a social loss of nearly 1,000 SKR per year. This, however, is an underestimate. The fact is that the average purchase made before the closure is not strictly comparable to the one made after the closure (Table III). Prior to the supposed closure 81% of all purchases were made at the local shop and only 5% by means of home deliveries, which is an expression of the preference of households: - to visit the local shop (with a chance of seeing the goods and making social contacts) rather than having the goods sent home - t o visit the local shop (with a chance of, inter alia, meeting people living in the locality) rather than to shop elsewhere. T A B L E III
Percentage Distribution of Methods of Purchasing Before and After the Supposed Closure of the Local Shop
Before
Visits to shop nearest shop while travelling to and from work at urban centre Home deliveries Total
After
95
57
81
16
5 9
25 16 5
43
100
100
446
Ernst Jonsson
After the closure purchases at the local shop are replaced by visits to shops elsewhere or home deliveries. Households therefore have to revert to shopping alternatives that are not valued as highly. In other words, the average method of purchasing after the closure is inferior in "quality" to the one that prevailed before the closure. The above estimate does not take into account the loss of welfare that households suffer due to this deterioration in "quality." If we are to try to include this loss, too, it will be necessary to start from the purchasing method chosen by the household (rather than the number of purchasing visits).
Computational Model B: Purchasing Methods Chosen by Households Before and After the Supposed Closure At the present time households visit their local shop an average of 2.4 times a week. If the shop is closed down they must then decide how these purchases (as well as 0.15 home deliveries per week from this shop) are to be made in future. What the households decide may then be interpreted as a result of their having weighed up a number of different options available to them. In each situation two alternative methods of purchasing (e.g., visits to a shop or home deliveries) are compared with each other. According to what emerged during the interviews the households do as a rule have a choice in the various respects that are relevant. For example, households that do not possess a car but are considering paying a visit to the shop may be able to travel in a car owned by a neighbour or acquaintance or by taxi. To illustrate the welfare loss that the closing down of a shop can inflict on households the following shows how they react to the four possible options.
Option No. I: Purchasing by means of home delivery or visit to a shop in nearest locality? Compared with home delivery a visit to a shop involves households in increased costs in time and travel. At present these extra costs amount to an average of 2.50 SKR per shop visit, that is to say, 3.40 SKR per shop visit less 0.90 SKR per home delivery. The majority of households are willing - prior to the closure of the shop to these extra costs in order to benefit from the advantages of the pay shop visit. If the local shop is closed down the extra cost of visiting the nearest shop (that is to say, in a neighbouring locality) increases to 19.50 SKR per visit. According to those interviewed, of the 2.4 visits per week to the local shop 0.65 are then replaced by home deliveries from the shop in the neighbouring locality. The households'
447
Costs and Benefits of a Shop Closure
per shop visit
C~ '2
19.5
. . . . . . . . . . . . . . . . . . . . . . .
ti:::=IJ;;.J; I .'.'.'
i'::q I.:'.' c.:" i"
1:;)2 I;;:'[ d."
J •
." :
":
::::.":7:[:".;. I:'.'::'.:.'i ".) I'. : ~:.'. ".':.". 2.s . . . . . . . . . . . . . . . . . . . . . . . . .
~. ~ :; : =..~.= ; " ' : " :'" " D
': ;.': : 2:'.: :.!:';', o. 7 . . . . . . . . . . . . . . . . . . . . . . . . . . 0.5
2 L ......................
~%.v. y:7..7 - T'-.7 B
b.'_..:_,-:: ..v'.'_.''.J
1.75 After
2.40 Before
D,Number o f visits to nearest shop
Fig. 3. E x t r a cost o f visiting the nearest shop c o m p a r e d with h o m e delivery, and n u m b e r o f visits, before and after closure o f the local shop. In S K R .
willingness to pay (= evaluation of the advantages of shop visits) ~s not sufficient to make up for the extra increased cost as regards the visits that are no longer paid. At present, it is since the extra cost of visiting the nearest shop (= the local one) is a good deal lower. Consequently, the shop visits that would be replaced by home deliveries in the event of closure could be expected to be accompanied by an annual loss of welfare to the household of: (Eq. 3)
52 X 0.65 (
2.50 + 19.50 2 - 2.50 willingness to pay per shop visit
) ~ 290 SKR
extra cost per shop visit borne by household
The social loss occasioned by visits not taking place is even greater, and is in fact equal to the dotted area in Figure 3. From the extra cost per shop visit borne by the household one must in fact deduct the extra cost incurred by the shopkeeper who has to arrange for home delivery, compared with the cost of shop visits. If home deliveries are utilized fully (say, 20 deliveries per round) this extra cost is estimated at 2 SKR per home delivery from the local shop. Including an adjustment for the social cost ot transport, one arrives at an estimate of the extra social cost of one visit to the local shop (compared with home delivery) of 0.7 SKR. However, the 0.65 visits
448
Ernst Jonsson
to the local shop are replaced not by home deliveries from that shop but by deliveries from the neighbouring locality, thus the cost of home deliveries is 1.20 SKR higher than would be incurred if deliveries were made from the local one. Thus, the social cost of a visit to the local shop (= 3.50 SKR) will be 0.50 (= 0.70 - 1..20) SKR lower than the corresponding cost of a home delivery from the shop in the neighbouring locality (4 SKR). The 0.65 visits per household per week are replaced, in other words, by a dearer shopping method which, furthermore, does not provide the advantages of a visit to a shop. This change leads to a social loss per household and year of: (Eq. 4)
52 × 0.65 (2.50 +2 19.50 + 0.50 willingness to pay per visit to shop
) ~ 390 SKR
reduced social cost per visit to shop
However, this estimate does not quite cover the total loss. The fact is that the visit to the nearest shop, on which Figure 3 (and also Figures 4-6) is based, is not identically the same before (-- to the local shop) and after (= visit to a shop in a neighbouring locality) the closure. The households in Vormsele have also demonstrated by their actions that they value a visit to their local shop more than they value a visit to a shop in a neighbouring locality. For one thing, a shop in a neighbouring locality cannot function to the same extent as a meeting place for the inhabitants as does a shop in their own locality. So, the household seems to have greater willingness to pay (e.g., the area ACDE in Figure 3) in order to visit its local shop than it does to 'visit a shop elsewhere (= area ABDE). The social loss sustained due to the 0.65 missing shop visits is therefore underestimated to a corresponding extent (= area BCD). As it has not been possible to calculate the whole of the willingness to pay visits to the local shop it is not possible to indicate the size of this underestimate. Option No. 2. Visit to shop in nearest locality or while travelling to and from work. If the local shop is closed down the average household will replace 0.65 visits to the shop by home deliveries, but then it will have to decide what is to be done about the remaining (that is to say, 2.40 - 0.65 = 1.75) visits which are at present made to the local shop. One possibility following the closure is to visit the shop in the nearest neighbouring locality. Compared with, for instance, purchases made while travelling to and from work this means:
449
Costs and Benefits of a Shop Closure
on the one hand, extra costs in time and travel - on the other hand, advantages (such as, less stress, more chances of social contacts). -
Extra cost per shop visit 19.o . . . . . . . . . . . . . . . . . . . .
-I
i):
g-,')
I');!
i%','.* I J2 -':'
vb:.'.',
i:.:..:;1 I :: "j 5: ============================= lo)
1 45 After
t 75 Numb~l~rof vlsits Before to nearestshop
Fig. 4. Extra cost of a visit to nearest shop compared with a visit made on the way to or from work, and n u m b e r o f visits, before and after closure o f the local shop. In SKR.
In the hypothetical situation without a local shop the privatelyborne extra cost of visiting the nearest shop (in relation to purchases made on the way to or from work) will rise from 2 to 19 SKR per visit. This means that shop visits per household will fall by 0.3 per week in favour of purchases made on the way to or from work. In relation to the present situation, where the willingness to pay for these shop visits is greater than the corresponding extra cost, this represents an annual loss of welfare per household of: (Eq. 5)
52 X 0.3 ( -2 . 0 +2 19.0 willingness to pay per visit
2.0
) ~
133 SKR
extra cost to household per visit
The extra cost (more precisely the transport cost) for a shop visit is, however, greater from the point of view of society than it is from that of the household (2.10 as against 2.00 SKR), which reduces the social loss to 130 SKR per year (= dotted area in Figure 4). For the same reason as the one mentioned in the previous section the loss is to some extent an underestimate.
Option No. 3. Visit to shop in nearest locality or in urban centre? Compared with a shop in one's own locality or in a nearby one,
450
Ernst Jonsson
shops in an urban centre may, on the one hand, have a larger and better range of goods. On the other hand, the private cost in travel and time of visiting a shop in an urban centre is 57 SKR higher than that of a corresponding visit to the nearest shop (that is, in one's own locality) since the average price level does not seem to differ between urban and rural shops (National Swedish Price and Cartel Office, 1982). If the local shop is closed down this extra private cost will fall to 40 SKR, and this leads to a slight increase in shop visits to the urban centre. In relation to the present situation the extra social cost (= 66 SKR) is, however, higher than the willingness to pay ((40.5 + 57.5)/2 = 49 SKR) for these additional (= 0.05 per household and week) shop visits to the urban centre. If the local shop is closed down the difference represents a social loss (dotted area in Figure 5) of: (Eq. 6)
52 X 0.05 (65.9
- 49
extra social cost per visit to urban centre
) ~ 45 SKR
willingness to pay for visit to shop in urban centre
Extra cost per shop visit 65.9 . . . . . . . . . . .
57.5
n%r.n i...', I:'"' ff..'.l
.......... -';::I Ii::l ":i
40.5 . . . . . . . . . .
11
0.25 0.30 Before After
N u m b e ~ o f visits to shop in urban centre
Fig. 5. Extra cost of a shop visit in urban centre compared with a visit to nearest shop, and number of visits to urban centre, before and after closure of a local shop. In SKR.
Option No. 4: How many o f the remaining visits to the nearest shop will be paid if the local one is closed down? Of the 2.4 visits per week that households at present - on the average - pay to the local shop 1.0 visit will therefore be replaced by other methods of
451
Costs and Benefits of a Shop Closure
purchasing in the event of its closure (0.65 by home deliveries, 0.30 by purchases made on the way to or from work, and 0.05 by purchases made at the urban centre). In this hypothetical chain of successive options the household must decide how many of the remaining 1.4 purchases per week must in fact be made if the local shop is closed down. The nearest shops that come into the picture for shopping visits are in Ruskmisk, Rusksele, and B]6rksele (all about 15 km from the local one) and a visit to one of these in terms of time and travel will cost about 20 SKR compared with 3.4 SKR per visit to the local shop. Thus, if the local shop is closed the cost of visiting the nearest shop will be six times as great as before. Apart from the shop visits that are replaced by other purchases the number of visits to the nearest shop will decrease by 1.1 per week. To be sure, households will be relieved of the time and travel costs they now incur on these visits to their local shop. If households are allowed to choose for themselves, which they can do if their local shop still exists, they will however prefer to make these 1.1 visits per week to the shop. The fact is that they value these visits (in terms of willingness to pay) higher than the corresponding costs in time and on travel. If the local shop is closed down the difference can therefore be regarded as a measure of welfare loss (= dotted area minus the tiny dark area in Figure 6) they suffer as a result of being unable to make the 1.1 visits per week to that shop: (Eq. 7)
52 X 1.1 (
3.4 + 20.4 2 - 3.5 willingness to pay for a visit to the nearest shop
) ~ 480 SKR
social cost of visit to the nearest shop
As Figure 6 shows the household continues to make 0.3 visits per week to the shop that lies nearest following the closing down of the local shop. The social cost of each such visit has, however, become 6.5 times as great as before (22.5 as against 3.5 SKR), and this means a social loss of nearly 300 SKR per household per year (= shaded area in Figure 6): (Eq. 8)
52 X 0.3 (22.5 - 3.5) ,~ 295 SKR
The Total Loss Suffered by Households At present the 50 or more households in Vormsele make an average of 2.95 purchases of everyday commodities per week and of these
452
Ernst Jonsson Cost per shop visit
22,5
1
~:..?'; j.-.!.::...:.'-,.:..
~,~ii!]ii~i!i~~:-'..:.':" v.-:. ~.(:.' ".:. 3.4
"-"
I .................
1 0,3 After
1,4 Before
Numbe~of visits to nearest shop
Fig. 6. Cost of a number of visits per household per week to the nearest shop, and number of visits, before and after closure of the local shop. In S KR.
2.4 involve a visit to the shop itself. If the local shop is closed down then: - 1.0 shop visits will be replaced by home deliveries, by purchases made going to or from work, or at the urban centre - 1.1 shop visits will no longer be made - 0.3 visits will be made to the nearest shop (mainly one of the three which are 15 km away in neighbouring localities). The social loss that each of these changes means for each household has been estimated in Figures 3 - 6 and Equations 4 - 8 . Overall, the changes imply a yearly loss of 1,340 SKR per household, though from this sum a deduction of 10 SKR must be made for the decrease in the cost of the 0.15 home deliveries per week already being made. As a result of the sharp increase in the number of home deliveries the presumed shop closure will involve, it will in fact be possible for the shop in the neighbouring locality to make the maximum use of its capacity. On the other hand, the costs now incurred in visits to the shop at the urban centre or on the way to and from work will not be affected by the closure of the local shop. Thus the deterioration that households will be subjected to can be estimated to correspond to a social loss of 1,330 SKR per year. Of this loss: 1,045 SKR represent the welfare loss suffered by households as a result of the fact that visits to the local shop are no longer made (480 SKR) or are replaced by some other mode of purchasing (565 SKR); -
453
Costs and Benefits of a Shop Closure
TABLE IV Yearly Social Loss per Household in the Event of the Closure of the Local Shop (SKR) Number o f purchases per week 1. Visits to nearest shop replaced by - home deliveries from shop in nearest locality - shop visits on way to and from work - shop visits at urban centre
Model B
Model A a
0.65
390
300
0.30
130
140
0.05
45
0.30
295
140
0.15
-10
70 70
565
25
465
2. Remaining visits to nearest shop Remaining home deliveries Remaining visits on way to and from work Remaining shop visits at urban centre
0.15
0
0.25
0
3. Visits no longer made to local shop
1.10
480
110
Total
2.95
1330
970
285
115
395
a Calculated as " N u m b e r o f purchases per week" × 52 × "Increase in social cost per purchase" (according to Equation 2 = 18.0 -- 9.1). The last entry emanates from Equation 1.
- nearly 300 SKR represent increased costs incurred in visits to the nearest remaining shop. Compared with this result, which is provided by Computational Model B, Model A underestimates the social loss suffered by households (Table IV). The underestimate is related entirely to the 1.1 visits to the local shop that are no longer made if the shop is closed (corresponding to a loss of 480 SKR as against 110 SKR). The fact is that Model A assumes that the 1.1 visits that are no longer made take the form of the average modes of purchasing, though in reality the visits no longer made are the 1.1 visits to the nearest shop, that is to say, the local one. The social cost incurred in time and travel in connection with these shop visits is a good deal lower than that incurred in the average mode of purchasing (3.5 as against 9.1 SKR). It is moreover this method of shopping that households value most highly.
454
Ernst Jonsson
As regards the modes of purchasing that remain after the closure or are replaced by other modes, the result obtained in Model A (a loss of 465 + 395 = 860 SKR) does not, on the other hand, differ very much from that obtained from Model B (565 + 285 = 850 SKR). As Model A is based on the average mode of purchasing, however, the result does not give a true picture of each one of the various modes of purchasing. As is seen from Table IV, on the one hand Model A underestimates the cost increase incurred in shop visits that remain after the closure. On the other hand, Model A erroneously assumes that increased costs are incurred in home deliveries and shop visits made in the urban centre or on the way to and from work.
Benefit Minus Costs The hypothetical closure will on one hand inflict a social loss of 72,000 SKR on all the households (incl. tourists) in the course of a year. On the other hand, however, closure of the shop would release capital resources (= shop premises, furniture, fittings, and stock) worth 90,000 SKR. Given a real yield of 8% this represents an annual income of just over 7,000 SKR. Even if the work formerly carried out by the closed-down shop can be done more efficiently by competing, larger shops, the savings in the labour force do not represent a real saving. The fact is that unless these resources can be transferred to regions with a permanent shortage of labour they have no alternative value. The working time that possibly will be saved by a shop closure can, at most, be given the same value as non-working time. If the shop studied here is closed down the social loss can be estimated at 65,000 SKR per year, though as regards the typical subsidized shop with twice the net turnover (800,000 SKR per year) the annual loss would be between 130,000 and 140,000 SKR. Thanks to the investment support provided by the government, the shop at Vormsele was reopened nearly five years ago, and by this time society has earned nearly four SKR for each one invested in order to put the business back on its feet. The longer the shop can remain open the greater this return will be.
Sensitivity Analysis Computational order to assess local shop. For households are
Model B makes use of a step-by-step procedure in the households' willingness to pay for keeping the this purpose it does not matter in which order the faced with the four various options, and so the
Costs and Benefits o f a Shop Closure
455
order in which these latter are presented is quite arbitrary as regards the result. The calculation presupposes that the shops that take over the business of the local shop in the event of its closure are in position to undertake relatively large-scale home deliveries of everyday goods. If this is not the case then home deliveries may be replaced by the households themselves visiting the shop in the nearest adjacent locality, and in that case the yearly loss would rise by 20%. In the calculation a comparison is made between an immediate closure of a local shop and the alternative in which it remains open for another year. If instead the alternative is remaining open for five years the yearly loss falls by 3%.
GENERAL APPLICABILITY OF THE METHOD
There is no abundance of cases of cost-benefit analysis applied to consumer policy issues. Among these attempts are those that are directed at facilitating purchases made by consumers, and here the cost incurred in travel and the time spent travelling, searching, and purchasing can be regarded as the price the consumer pays for a certain mode of purchasing. For this reason it is possible to transfer the methodology used when deriving the willingness to pay that is traditionally applied in cost-benefit analysis within the transport sphere. This methodology must, however, be modified in case the measure studied also affects the pattern of purchasing adopted by consumers. In that case the derivation of the willingness to pay can be based on their actual choices among the options available to them in various situations. In this study the cost-benefit methodology has been applied to a situation where the number of options has been reduced (that is to say, by the option of making purchases at their own local shop). The methodology is just as useful for estimating the willingness to pay, for example, for a measure that increases the number of options (by, for example, the establishment of a hypermarket). REFERENCES Bohm, P. (1972). Estimating demand for public goods: An experiment. European Economic Review, 3, 111-130. Harrison, A. J., & Quarmby, D. A. (1972). The value of time. In: R. Layard (Ed.), Costbenefit analysis, pp. 173-208. Harmondsworth: Penguin. Jgnsson, B. (1976). Cost-benefit analysis in public health and medical care. Lund: Lund Economic Studies.
456
Ernst Jonsson
Mishan, E. J. (1972). Cost-benefit analysis. London: Allen & Unwin. National Swedish Price and Cartel Office (1982). Prisniv~,er i dagligvaruhandeln, Ume~ oktober 1981 (Price levels in shops for everyday commodities, Ume~ October 1981). Stockholm: National Swedish Price and Cartel Office. SPK:s utredningsserie 1982:3. Pearce, D. W. (1971). Cost-benefit analysis. London: Macmillan.
ZUSAMMENFASSUNG
Lohnt es sich, Einzelhandelsgeschhfte in diinnbesiedelten Regionen aufrecht zu erhalten? Eine Kosten-Nutzen-Anatyse anhand eines hypothetischen schwedischen Falles. Die SchlieBung des einzigen Gesch~iftes in einer entlegenen Ortschaft ffihrt f(ir die Haushalte zu h~heren Einkaufskosten und zu anderen WohlfahrtseinbuBen. Diesen Kosten stehen m~Jgliche Gewinne durch das Freiwerden yon Arbeit, von Geb~iuden und Einrichtungen und yon Lagerbest~inden gegen/Jber. Ob die wirtschaftliche Wirkung einer Gesch?ftsschlieBung ftir die Gesellschaft positiv oder negativist, h~ingt vor allem ab (a) yon der Entfernung zum n~tchst erreichbaren Laden, (b) v o n d e r Zahl der betroffenen Haushalte und (c) vom Nutzen der Alternativ-Verwendung der freiwerdenden Ressourcen. Um soziale Kosten zu vermeiden, gibt der schwedische Staat finanzielle Hilfen, damit L~den in entlegenen Gebieten nicht geschlossen werden. Der vorliegende Beitrag versucht, die Wirkungen dieser staatlichen Hilfen als die vermiedenen sozialen Kosten einer Gesch~iftsschliegung zu quantifizieren, und geht aus v o n d e r hypothetischen SchlieBung eines realen Gesch~iftes. Ausgew~ihlt wurde das Geschhft in der nordschwedischen Ortschaft Vormsele, das von Dezember 1976 bis M~irz 1978 geschlossen war und dann mit staatlicher Hilfe wieder ge/Sffnet wurde. Diese quasi-experimentelle Bedingung macht das Gesch~ift besonders geeignet f~ir die Modell-Rechnung. Es ist ca. 15 km vom n~ichsten Gesch/ift und ca. 35 km v o n d e r nhchsten st~idtischen Einkaufsm6glichkeit entfernt und entspricht damit ziemlich genau dem Durchschnitt der staatlich unterst/itzten Geschhfte. Die Berechnungen, fiber die der Beitrag berichtet, ffihren zu dem Hauptergebnis, dab die SchlieBung eines typischen staatlich unterstfitzten Gesch~iftes mit einem durchschnittlichen jhhrlichen Umsatz von 800.000 SKR zu einem jhhrlichen gesellschaftlichen Verlust von 130.000-140.000 SKR fi2hren wiirde. Der Gewinn, den die staatliche Unterst~itzung des Geschhftes in Vormsele in den bisherigen fiinf Jahren gebracht hat, ist viermal so hoch wie der Unterstiitzungsbetrag. Der Gewinn wird um so h6her werden, je lhnger das Geschhft geiSffnet bleiben kann.
THE AUTHOR Ernst Jonsson is an Associate Professor at the Department of Business Administration, University of Stockholm, S-106 91 Stockholm, Sweden.