Content uploaded by Fernando Rodrigo Sauco
Author content
All content in this area was uploaded by Fernando Rodrigo Sauco on Mar 02, 2015
Content may be subject to copyright.
Cross-Border Shopping: A Survey
Andrés Leal & Julio López-Laborda &
Fernando Rodrigo
Published online: 21 February 2010
#
International Atlantic Economic Society 2010
Abstract The aim of this paper is to undertake a revie w of the most important
literature on the phenomenon of fiscally induced cross-border shopping. Following
the presentation of the principal theoretical models, the study concentrates on
applied literature. Firstly, the elements common to the diverse applications are
described, and then, a detailed analysis of the research undertaken into cross-border
shopping for alcoholic drinks, tobacco, fuel, and lotteries is provided, concluding
with a reference to the interaction between cross-border purchases and those effected
over the internet. The results achieved by the empirical research coincide and
support the principal result of the theoretical literature: the tax differentials between
neighboring territories induce consumers to purchase in the territory where taxation
is lower, on the condition that the tax saving compensates for the transport costs
associated with the travel made by the purchaser in order to take advantage of the
lower taxation.
Keywords Cross-border shopping
.
Tax differences
JEL H31
.
H32
.
H73
.
E62
Int Adv Econ Res (2010) 16:135–148
DOI 10.1007/s11294-010-9258-z
A. Leal
Department of Economics, Finance and Employment, Government of Aragon, Plaza de los Sitios, 7,
50.001 Zaragoza, Spain
e-mail: aleal@aragon.es
J. López-Laborda
:
F. Rodrigo (*)
Department of Public Economics, University of Zaragoza, Gran Vía, 2, 50.005 Zaragoza, Spain
e-mail: frodrigo@unizar.es
J. López-Laborda
e-mail: julio.lopez@unizar.es
Introduction
The differences in the taxation of the same good or service between neighboring
countries, neighboring regions, or municipalities in the same country encourage
consumers to travel to the jurisdiction where taxation is lower to acquire that good or
service, as long as the tax saving compensates for the costs of traveling from one
jurisdiction to another. This behavior has consequences for tax revenue collection
and economic activity in the affected jurisdictions and for the welfare of individuals
in those territories.
This phenomenon, known as cross-border shopping, has been widely recognized
and analyzed in the literature since at least the 1930s. The existence of cross-border
shopping has been shown for various time periods, spatial areas (countries, regions,
local authorities), taxes (general on sales or on specific products), and goods and
services (food, clothing, alcohol, tobacco, fuel, lotteries, etc.). The importance of
cross-border shopping increases as two trends in the world economy intensify. The
first is the globalization of the economy, which substantially reduces the transaction
costs associated with trade relations between territories. The second trend is the
decentralization of tax-raising and expenditure powers towards intermediate level s of
government within the same country, which significantly increases the number of
jurisdictions with differentiated taxation of the same goods or services.
The aim o f the present paper is to present a systematic review of the most
important literature on cross-borde r shopping from a dual perspective, theoretical
and applied. The theoretical works are presented in the second section of the paper
and the applied works in the third. In order to reduce the review to a reasonable size,
we shall concentrate exclusively on research directly related to cross-border
shopping, leaving aside other subjects of undeniable importance to which that
phenomenon is directly related, such as tax competition or tax incidence. Each of
these topics could be the subject of a survey in itself, one more extensive than that
offered here. Finally, we are not going to consider those border-crossing transactions
that involve no payment of tax. They are referred to as tax-not-paid (TNP)
transactions, being a well-known example of these the VAT carousel fraud.
1
Theoretical Works
On the theoretical plane, the standard work is by Kanbur and Keen (1993)
2
.
Although it must be noted that, as we shall see in the following section, some
previously applied research had already been firmly grounded on standard models of
economic theory (such as, for example, Fisher, 1980).
The partial equilibrium model supplied by Kanbur and Keen (1993 ) considers an
open economy with a single good and comprised of two bordering countries—
national and foreign—that differ in population size. The objective of the govern-
ments is to maximize revenue. The consumer will purchase the good if the price is
1
TNP transactions are analysed by Keen (2002).
2
Which the authors themselves relate to certain classic literature in fiscal federalism, such as Gordon
(1983), Mintz and Tulkens (1986) and Lockwood (1993).
136 A. Leal et al.
inferior or equal to its reservation price: this is v for the national resident and V for
the foreigner. The price for the producer of the commodity is constant and the same
in both countries, and thus, the consumer price can be identified with the tax paid in
each territory; t in the country of residence and T abroad. Each store must charge the
tax rate of the country in which it is located.
In this model, the national consumer faces the alternative of buying the good,
mentioned above, in his or her own country or of traveling to the border and
purchasing it in the neighboring country, incurring a cost of δ>0 per unit of distance,
s. There are two necessary and sufficient conditions for cross-border shopping to
occur: the surplus that the consumer enjoys by buying abroad exceeds that of doing
so in the national market, that is to say v T d:s > v t, and this surplus is non-
negative: v T d:s > 0.
In a closed economy, cross-border shopping does not exist and each government
fixes its taxation at the level which coincides with the reservation price of its
residents. When moving from a closed economy to an open one, in a non-
cooperative model, the smaller country will reduce its tax rates, which will cause the
following: a reduction in the tax revenue collection of the larger country (caused by
cross-border shopping); an increase in that of the smaller country, if the difference in
their size is sufficiently high; and a reduction of the aggregate revenue. In the Nash
equilibrium, the tax rate of the smaller country will be lower and the per capita
revenue collection (but not absolute revenue) greater than the larger country.
Kanbur and Keen (1993) identified two strategies to achieve increases in welfare:
an increase in trans port costs (which permits taxation and revenue to be increased in
the two countries, without modifying the volume of cross-border shoppi ng) and the
establishment of a minimum common tax rate, being the best strategic response of
the larger country to ensure a sufficient level of cross-border trade from which the
smaller country would also benefit. On the basis of the work of Kanbur and Keen
(1993), the literature has been extended in many directions. While some studies
maintain the emphasis on the difference in the population size of territories (Trandel
1994; Wang 1999), others assume that territories differ in their geographical size
(Ohsawa 1999; Nielsen 2001, 2002). Some authors introduce diverse transaction
costs (Scharf 1999; Nielsen 2001) or differences in the preferences for public goods
(Haufler 1996; Nielsen 2002). In general, the results obtained by such research
coincide with those of Kanbur and Keen (1993 ). A final group of studies is
concerned with deriving optimal taxation rules for economies exposed to cross-
border shopping, in various scenarios (Scharf 1999; Christiansen 1994, 2003). We
present the most important results from this literature below.
Trandel (1994 ) confirms, for diverse hypotheses regarding the behavior of
governments and firms, that the more densely populated country imposes a higher
tax rate. Wang (1999) broadens the preceding models, since he considers that the
more highly-populated region behaves as a Stackelberg leader.
3
The results show
that the tax rates and the revenue of both regions are greater than those which would
correspond in the Nash equilibrium, and the reforms aimed at fiscal harmonization or
rules of minimum rates are damaging for the smal ler region, while they benefit the
larger one.
3
See also Hvidt and Nielsen (2001).
Cross-Border Shopping 137
Ohsawa (1999) introduces two modifications to the approach of Kanbur and Keen
(1993). Firstly, the author extends the model to more than two countries to understand
how the relative position of the countries affects equilibrium. Secondly, he assumes
that the countries are of different geographical sizes. Both characteristics—positions
and sizes of the countries—create differences in the market power, which in turn
produces differences in tax rates and government revenue in the Nash equilibrium.
The results confirm those of Kanbur and Keen (1993), since they show how small
countries choose, in the Nash equilibrium, a lower tax rate than large countries, even
though they collect more than its size-proportional share of total revenue. Ohsawa
(1999) also proves that, if countries are of the same size, the equilibrium tax rates
diminish gradually from the peripheral countries towards the central one.
Nielsen (2001) also locates the asymmetry between countries in their geograph-
ical extension. The results achieved confirm once more those obtained by Kanbur
and Keen (1993). Nielsen (2001) incorporates two additional extensions to the
preceding research. Firstly, he shows that the consideration to the transport costs for
commodities (as well as persons) leads to higher taxes and, in the smaller country, to
the existence of economic rents for sellers close to the border. The attempt to tax
such rents will nevertheless reduce revenue collection in that country, due to the
strategic reaction of the larger country. Secondly, Nielsen (2001) proves that the
existence of border audits raises taxes in both countries and may increase the volume
of cross-border trade.
Haufler (1996) introduces differences in the preferences for public goods. The
author analyz es strategic taxation in a model with trade in on ly one private good,
which is simultaneously imported by the consumers residing in a country with a high
tax level and exported by its producers. If the differences between countries in their
preferences for public goods are small, Haufler (1996) demonstrates the existence of
the Nash equilibrium, in which the country with the greater preference for public
goods has a higher tax rate. Subsequently, Haufler (1996 ) considers the two tax
coordination measures exami ned by Kanbur and Keen (1993): a minimum tax rate
and a coordinated increase in the transport costs, obtaining less conclusive results.
Tax coordination in general benefits countries with high taxes, while the welfare of
countries with lower taxes will be reduced if tax coordination reduces the tax rate in
the other country. This last case may occur when tax competition in the initial
equilibrium is moderate, due to the relatively high costs of cross-border shopping, or
when the elasticity of substitution between private and public consumption is low in
the high-tax country.
Following the methodology of his previous work, Nielsen (2002) show s that it is
possible to generate cross-border shopping from small countries towards larger ones,
as long as the marginal cost of public funds in the smaller country sufficiently
exceeds that of the bigger one, whether because the citizens of the smaller country
value public goods more highly than the citizens of the larger country, or because the
taxes other than those on sales are more distorting in the smaller country. The wedge
created in the marginal cost of public funds justifies, in his opinion, the different
examples existing in the real world of cross-border shopping from small countries to
large ones, such as from Canada to the United States or from Denm ark to Germany.
As explained above, a final group of studies attempts to derive optimal taxation
rules in diverse scenarios. Scharf (1999) proposes a model in which she considers
138 A. Leal et al.
the existence of transaction and storage costs. Consumers incur fixed transaction and
transport costs to access in the foreign market a perfect substitute good for the
national good. The results show that the size of the optimal tax is inversely
proportional to the volume of national transactions. This result permits Scharf (1999)
to state that when there exist increasing returns to scale in tax avoidance with respect
to the quantities traded, then smaller transactions must be taxed more heavily than
larger ones.
Christiansen ( 1994) analyzes the effects of market structure upon cross-border
shopping. In a competitive market, the traditional inverse elasticity rule (in this case,
of domestic demand) represents a valid characterization of the optimal tax rate.
When domestic supply comes from a foreign monopoly, a modified inverse elasticity
rule is applied, due to the possibility that part of the tax burden may be shifted onto
the owners of the firm. Lastly, the most predictable response to a rise in prices or
taxes abroad will be an increase in domestic taxes.
Christiansen (2003) also inves tigates the implications of cross-border shopping
for the tax structure of a high-tax country which is exposed to the purchas es of its
residents in a neighboring country. Firstly, the paper shows that taxes on products
that generate externalities (such as excise duties on cigarettes, alcohol, or fuel) must
be fixed at a level lower than that determined by Pigou’s rule, in order to compensate
for the inefficiencies derived from cross-border shopping. With regard to taxes
established to collect revenue (such as general taxes on sales), Christiansen (2003)
explains that the reduction of the rates imposed upon goods subject to cross-border
shopping may not be an advisable policy, since it will produce distortions in the
consumption of the remaining goods subject to taxation. Second, when it is
considered that an individual may acquire severa l products at the same time in the
neighboring country, Christiansen (2003) concludes that the most efficient way of
containing cross-border shopping may be to reduce the tax on the products
purchased both inside and outside the country.
Applied Research
General Aspects
Although, according to Mikesell (1970), interest in cross-border shopping induced
from tax differences could already be detected in the United States from the mid
1930s onwards, the first applied research was performed between the 1950s and
1970s (Maliet 1955; McAllister 1961; Hamovitch 1966; Levin 1966; Mikesell 1970,
1971), and it established the fundamental methodological aspects which today are
prevalent in this literature.
In its most general form, the following demand function is used:
S
r
¼ SP
r
; P
r
; T
r
; T
r
; Y
r
; C
r
; XðÞ ð1Þ
In general, S
r
are sales in territory r (being r a country, region or locality) of the
product or products under analysis, although some studies use consumption as the
endogenous variable. As explanatory variables, P
r
and P
r
, T
r
and T
r
are,
respectively, the price s and taxes of the products analyzed, in territory r and in its
Cross-Border Shopping 139
neighboring territories, r, Y
r
is the income in territory r, C
r
is the travel cost from this
territory to its neighboring ones, and X is a set of additional control variables. The
variables related to transport costs were not explicitly incorporated into the empirical
models until the work performed by Fox (1986), though since then they have been
habitual in them. With regard to the additional control variables, their inclusion is
determined by the specific object ive sought in each application: the number of
Fridays in the month, to explain the sales of alcohol (Asplund et al. 2007); the stock
of vehicles, to explain the sales of gasoline (Banfi et al. 2005); the expenditure on
advertising of the tobacco companies, to explain the sales of tobacco (Coats 1995);
and so on.
Usually, the specifications take a multiplicative form, which favors subsequent
logarithmic transformation and a direct interpretation of the estimated coefficients in
terms of elasticity. An illustration of this is provided by the work of Walsh and Jones
(1988) which, despite being very simple, is also quite complete. These authors
analyze the effect derived from a reduction of 3%, by one point annually, of the sales
tax rate in the state of West Virginia in the 3 year period, 1980–1982. To do this,
Walsh and Jones (1988) formulate the following multiplicative demand model:
S
it
¼ A
i
Y
a
it
P
b
it
C
c
i
ð2Þ
with which they attempt to explain the impact upon per capita sales of taxed goods,
S, of variables such as income per capita, Y, the after-tax price of goods in the county
relative to that available in other adjacent locations, P ¼
p 1þTðÞ
p
a
1þT
a
ðÞ
, and the transport
cost associated with the journey to purchase in other locations, C, measured as the
average distance (in miles) of the residents of West Virginia with regard to the
nearest comm ercial centre in an adjacent state. A is a multiplicative factor. Finally, i
denotes the county and t the time period.
Walsh and Jones (1988) estimate Eq. 2 in logarithmic form, using a data panel of
46 counties between 1979 and 1984:
ln S
it
¼ ln A
i
þ a ln Y
it
þ b ln P
it
þ c ln C
i
þ u
it
ð3Þ
Their results suggest that the residents of the counties that border West Virginia
choose to purchase outside their state, in order to avoid the higher taxation which
they would bear in their county of residence and display the opposite behaviour
when faced with reductions in their own county taxes: a reduction in the tax rate of
one per cent means an increase in sales of 5.9%. Walsh and Jones (1988 ) also
demonstrate the absence of cross-border shopping in areas distant from the border.
The results obtained by Walsh and Jones (1988) are a perfectly representative
example of those obtained unanimously by the literature on cross-border shopping:—
the tax differentials between bordering territories induce consumers to purchase in
those territories where taxation is lower, as long as the tax saving compensates for the
transport costs associated with the journey of the purchaser to take advantage of the
lower taxation. The magnitude of the impact of tax differences on sales is no small
matter. Restricting ourselves, for the moment, to the general tax on sales and the
taxation on food and clothing in the United States (a country which has available
research which permits the understanding of the effects of its taxation on sales from
the early 1950s until today, for diverse territorial areas and taxed products), the
140 A. Leal et al.
estimations performed repeatedly reflect than an increase by one percentage point in
the rate of tax reduces sales in the territory affected by, approximately, 6%.
4
Naturally,
this reduction in sales is not incompatible with an increase in tax revenue collection,
due to the increase in the tax rate.
Alcohol, Cigarettes, Gasoline, Lotteries
Although the existence of the border effect—and its impact upon revenue—has been
demonstrated for various goods, time periods and geographical areas,
5
applied
research has concentrated on testing the existence of cross-border shopping for some
specific products: alcoholic drinks, cigarettes, gasoline and gambling. This section
examines in detail a study devoted to each of these products, with the aim of
emphasizing how the general model presented in the previous subsection is adapted
to each specific problem.
Asplund et al. (2007) estimate, following the methodology of Walsh and Jones
(1988), how the monthly sales of alcoho l in Sweden, q, respond to prices in
Denmark and Germany, P
F
, taking into account at the same time the sensitivity of
the distance to the border:
$
12
ln q½¼b
0
þ b
1
$
12
ln P
D
þ gd; DðÞ$
12
ln P
F
þ b$
12
X þ " ð4Þ
where P
D
is the domestic price, gd; DðÞ¼d
0
þ d
1
D þ d
2
D
2
þ d
3
D
3
is the elasticity
with regard to the prices abroad, which depends on the distance to the border, D, and
finally X is a vector which represents the varying seasonal behavior of consumption:
concretely, it lists the number of Fridays, the day on which most alcohol is
traditionally sold. The exercise is performed using a panel of 287 municipalities, for
the period betw een 1995 and 2004. The authors estimate Eq. 4 as a system of
seemingly unrelated equations (SUR).
Asplund et al. (2007) conclude that the coefficients are, as a whole, significant;
they display the expected signs and are plausible as regards magnitude. The results
suggest that elasticity in border municipalities with respect to foreign prices is
approximately 30%; this is reduced to 20% (10%) when the distance is widened to
150 (400) kilometers, thereby reflecting the decisive role of distance to the border in
the extension of cross-border arbitrage. Furthermore, the lower level of German
prices compared to Danish ones compensates for the great er cost of the journey to
Germany, both in terms of time and outlays. The extra distance associated with
buying in Germany instead of Denmark may be economically profitable for some
consumers, espec ially those who purchase large quant ities. The diversion of
consumption towards the other Swedish border, that of Finland, demonstrates the
4
See Hamovitch (1966), Levin (1966), Mikesell (1970), Fisher (1980), Fox (1986), Mikesell and Zorn
(1986) and Tosun and Skidmore (2007).
5
For example, Boisvert and Thirsk (1994) and Di Matteo and Di Matteo (1996) have empirically
demonstrated the influence which the Canadian Goods and Service Tax plays in the decision of consumers
in that country to travel to the United States border to purchase. Ferris (2000) shows that the greatest
proportion of revenue lost in Canada due to cross-border shopping is derived from professional or
organised smuggling. This is, in aggregate terms, three times higher than the decreases generated by
private cross-border shopping. Fitzgerald (1992) provides evidence of cross-border shopping in Ireland,
and Gordon and Nielsen (2001) do the same in the case of Denmark.
Cross-Border Shopping 141
negligible influence of distance upon the sales of wine and beer, while the sale of
spirits is highly sensitive to it. The elasticity in the Finnish border is greater than in
the Danish one, given the absence of direct costs in crossing the border.
Lastly, the authors estimate that the reduction in Danish taxes on alcohol carried
out in October 2003 may have produced a 2.2% reduction in the revenue collection
of Sweden, with more than a quarter of that effect concentrated on the localities
closest to the border.
The cross-border purchases of alcoholic drinks has also been tested by Crawford
and Tanner (1995) and Crawford et al. (1999) for the United Kingdom, by Smith
(1976) and Beard et al. (1997) for the United States and by Fleenor (1999) for
Canada and the United States.
There are a great number of studies concerning tobacco cross-border shopping,
especially in the United States, where smuggling of this product is a traditional
concern of the authorities (Fisher 2007: 394–5). The existence of cross-border
shopping has been demonstrated empirically by, among others, Wertz (1971),
Warner (1982), Thursby et al. (1991), Coats ( 1995), Saba et al. (1995), Fleenor
(1998), Alamar et al. (2003), and more recently, by Chiou and Muehlegger (2008).
The results obtained by the literature consistently show that between 2% and 6% of
the cigarettes consumed in the United States are smuggled.
Using microdata for cigarette consumption in the period 1992–2002 in the
Metropolitan Statistical Areas (MSA) of the United States, Lovenheim (2008) develops
and estimates a demand model for cigarettes, which explicitly incorporates the decision
to purchase in a bordering state. The reduced form demand equation is as follows:
Π
0
þ Π
1
ln P
h
ðÞþΠ
2
ln P
h
ðÞln P
b
ðÞðÞþΠ
3
ln P
h
ðÞln P
b
ðÞðÞ
2
þ
þΠ
4
ln D
i
ðÞln P
h
ðÞln P
b
ðÞðÞþgX
i
ð5Þ
The dependent variable is alternatively the number of cigarettes smoked daily by
smokers, the rate of participation of smokers, and the number of cigarettes smoked
per day, including non-smokers. The explanatory variables of interest are the price
and the tax, in real terms, in the state of residence (denoted by subscript h) and in the
closest neighboring locality with a lower price or tax (subscript b) and the distance to
this locality (D). Age, gender, salary, marital status, race, educat ion, and employment
situation are considered as demographic variables (X). A time trend and fixed spatial
effects are also included.
The principal results obtained by Lovenheim (2008) are as follows. Firstly, the
price elasticity of the state of residence (that is, the percentage change in the
consumption of residents when the price in the state of residence changes by 1%) is,
on average, indistinguishable from zero. This means that, in the presence of price
differences between localities (and, therefore, of cross-border shopping), changes in
the price of tobacco have no effect upon the consumption of this product. By
contrast, and secondly, the total price elasticity (i.e. the percentage change in
consumption when all the prices change by 1%, and thus, the incentives to smuggle
are not modified ) is negative and of a considerable magni tude, but inelastic. In
accordance with this second result, state taxat ion of cigarettes would be a good
instrument to reduce cigarette consumption and to obtain public revenue, if
smuggling between states were eradicated.
142 A. Leal et al.
Lovenheim (2008) estimates that cross-border purchases slightly increase tobacco
consumption, and that between 13.1% and 25.1% of consumers in metropolitan
areas participate in cross-border shopping for cigarettes.
Finally, Lovenheim (2008) warns that the average results given above hide
considerable heterogeneity among states, derived from the geographical distribution
of the population. To give just two examples, in Washington, D.C. (which is three
miles from Virginia), 63.48% of consumers casually smuggle, and New Hampshire
(which is the New England state with the lowest taxes) doubles its sales of tobacco
due to smuggling.
With regard to cross-border fuelling, Banfi et al. (2005) estimated, using the data
panel technique, the impact of gasoline price differences between the border regions
of Switzerland and adjacent areas in Germany, Italy, and France have upon the
demand for this fuel in the Swiss border areas. They propose the following
specification:
ln G
it
¼ a
0
þ a
1
ln P
Swiss
ðÞ
it
þ a
2
ln
P
For
P
Swiss
it
þ a
3
ln
P
For
P
Swiss
it
hi
ln
N
For
N
Swiss
it
hi
þ
þa
4
ln
I
Swiss
N
Swiss
it
þ a
5
ln
I
For
N
For
it
þ a
6
ln N
Swiss
ðÞ
it
þ
þa
7
ln N
For
ðÞ
it
þ a
8
ln CommðÞ
it
þ a
9
ln CarsðÞ
it
þ m
it
ð6Þ
where G
it
represents the demand for gasoline in the Swiss border regions -defined as
those located less than five kilometres from Italy, Germany, or France- approximated
by the sales of gasoline of the three principal oil companies (SHELL, BP and
ESSO). (P
Swiss
)
it
and (P
For
)
it
are, respectively, the real prices of gasoline in the
border regions of Switzerland and in the foreign counties adjacent to them. (I
Swiss
)
it
and (I
For
)
it
,(N
Swiss
)
it
and (N
For
)
it
represent the income and population, respectively,
of the Swiss and foreign border regions. Finally, Comm
it
represents the daily
commuters who travel to Switzerland from the foreign countries’ border regions and
Cars
it
the stock of vehicles in each border region. The sub-index i denotes the three
Swiss border regions and t=1985,...,1997, the period under study.
The estimation permits the detection of an important cross-border effect since, in
the case of a reduction of 10% in the Swiss price of gasoline, demand in the Swiss
border areas would increase by almost 17.5%. Moreover, the authors also show with
their results that if a proposed tax on CO
2
emissions were finally approved in
Switzerland, this would eliminate a considerable part of car journeys caused by price
differentials.
Rietveld et al. (2001) test the existence of cross-border fuelling between The
Netherlands, Germany and Belgium; Doyle and Samphantharak (2008)and
Manuszak and Moul (2008) do so for diverse regions of the United States, while
Leal et al. (2009) undertake the same exercise among Spanish regions.
The last group of research we wish to concentrate on in this section is related to
the cross-border shopping produced by the taxation of gambling. Garrett and Marsh
(2002) undertake the first analysis in the United States, estimating the existence of
cross-border lottery shopp ing among the borde r counties of Kan sas and its
neighboring states, using 1998 data corresponding to the 105 Kans as counties.
Garrett and Marsh (2002) propose a regression of per capita sales of lottery tickets
in each county, using a dummy variable for each border state and a set of control
Cross-Border Shopping 143
variables: income, education, religious affiliation, ethnicity, tourist infrastructure,
number of miles of road in the county, urban population and the number of retail
lottery sales outlets available in the county. They also incl ude dummy variables
indicative of the existence of casinos and parimutuel racetracks.
Estimating by ordinary least squares and by maximum likelihood, to correct for
spatial correlation, Garrett and Marsh (2002) confirm the existence of cross-border
lottery shopping, and calculate that this phenomenon entails losses for Kansas of
approximately 10.5 million dollars or 5.4% of its lottery sales. Tosun and Skidmore
(2004) and Skidmore and Tosun (2008) have also provided empirical evidence of
cross-border lottery shopping.
Internet
The impressive development of trade over the Internet gives rise to very interesting
challenges for applied research, since it widens the list of options available to
purchasers. Internet purchasing is now an alternative to purchasing in the place of
residence or in a border locality. Goolsbee (2000), Alm and Melnik (2005), and
Goolsbee et al. (2007) analyze the effects of sales taxes upon the decision of an
individual to buy in his or her locality or over the Internet; they find a positive and
significant relationship between residence in a locality with high taxes and the
probability of purchasing via Internet. Goolsbee (2000) estimates tax elasticity in the
range of two to four, while Alm and Melnik (2005 ) calculate an elasticity of
approximately 0.5.
Ballard and Lee (2007) combine the analysis of the effects of taxes upon Internet
purchases with the analysis of cross-border shopping. To do this, and using data for
the United States from the Current Population Survey for the years 1997 and 2001,
they estimate a system with two probit equations, which includes a selection
equation for access to Internet and another equation for Internet purchases. The
specification of both equations coincides:
y
i
¼ b
0
þ b
1
HOMETAXPRICE
i
þ b
2
TAXRATIO
i
þ b
3
TAXBASE
i
þ
þb
4
LOGINCOME
i
þ b
5
FEMALE
i
þ b
6
WHITE
i
þ
þb
7
MARRIED
i
þ b
8
FEMALE
»
MARRIEDðÞ
i
þ
þb
9
WHITE
»
MARRIEDðÞ
i
þ b
10
FEMALE
»
WHITEðÞ
i
þ
þb
11
FEMALE
»
WHITE
»
MARRIEDðÞ
i
þ b
12
HIGHGRAD
i
þ
þb
13
COLLGRAD
i
þ b
14
PROGRAD
i
þ b
15
NUMCOMP
i
þ
þb
16
AGE15
i
þ b
17
AGE20
i
þ b
18
AGE30
i
þ b
19
AGE50
i
þ b
20
AGE60
i
þ
þb
21
D2001
i
þ b
22
CPICHANGE
i
þ COUNTYDUMMIES þ u
i
ð9Þ
The dependent variable y
i
is a binary variable which reflects, depending on the
equation estimated, whether the individual accesses Internet, and whether he or she
purchases via Internet.
Among the explanatory variables of a fiscal nature, HOMETAXPRICE represents
the sales tax rate in the county of residence; TAXRATIO reflects the ratio between the
domestic rate and the lower tax rate in the bordering counties; TAXBASE measures
the width of the taxable base of the state sales tax. LOGINCOME captures, in
logarithms, family income, NUMCOMP the number of computers in the household,
and CPICHANGE the percentage change in the Consumer Price Index. Additi onally,
144 A. Leal et al.
a set of binary variables are incorporated to reflect characteristics of demography
(FEMALE, WHITE and MARRIED), education (HIGHGRAD, COLLGRAD and
PROGRAD) and different ages (AGE(⋅)). Finally, D2001 is a dummy variable which
takes the value of one when the data correspond to 2001 and zero when they refer to
1997, and COUNTYDUMMIES controls for the existence of specific effects of the
counties.
Ballard and Lee (2007) obtain two basic results. Firstly, a resident in a county
with high tax rates is more likely to purchase via Internet than a resident in a county
with low rates. Secondly, the resident in a county which borders another one with a
lower tax rate or a narrower tax base is less likely to purchase via Internet, ceteris
paribus. For Ballard and Lee (2007), these results are consistent with the
interpretation that home-country shopping, cross-border shopping and Internet
shopping are substitutes.
Discussion of the Literature and Economic Policy Implications
In light of the wide-ranging literature reviewed in the previous sections, three
fundamental conclusions may be underlined. Firstly, the majority of theoretical
literature appears to support, from a variety of methodological approaches, the results
of Kanbur and Keen (1993), since they show how small countries select, in the Nash
equilibrium, a lower tax rate on commodities than large ones, although they collect a
percentage higher than that corresponding to their relative size. Secondly, the empirical
literature follows the same basic strategy, namely of specifying and estimating a demand
function whose essential arguments are the prices, taxes and income in the territories
affected. The studies differ in the selection of the control variables which, logically, are
adapted to the specific problem tackled in each application, and in the econometric
technique employed, although the use of panel data estimation is predominant. Thirdly,
the results achieved by empirical research clearly coincide and support the principal
result of theory: tax differentials between neighboring areas induce consumers to
purchase in that area where taxation is lower than, as long as the tax saving compensates
for the transport costs related to the journey made by the purchaser.
These results have obvious economic policy implications, which cannot be ignored
by governments when designing their tax policies. If a jurisdiction establishes a higher
tax on commodities than other jurisdictions (especially those areas closest to it), part of
the purchases which had previously been made in that jurisdiction will be diverted to
others, and this will firstly affect revenue collection in the jurisdiction affected, due to
this tax and to those which may, in turn, be charged upon those same purchases (such as,
excise duties and VAT). The importance of these effects must not be underestimated. In
the case of the United States, Manuszak and Moul (2008) have recently estimated that
the area of Chicago, the territory with the highest gasoline taxes, is losing approximately
40% of the tax-raising capacity that would exist if taxes were equal throughout the
region. In addition to the direct effects upon revenue collection, and as Fisher (2007)
notes, the reduction in purchases may affect retail sales activity, employment and house
prices in the jurisdiction and, consequently, tax revenue from property, income, and
sales. Finally, as Keen points out (2002), cross-border transactions will also involve
welfare effects.
Cross-Border Shopping 145
As Crawford et al. (2008) state, the policy response appropriate to cross-border
shopping will depend on the evaluation made of government behavior. If it is
believed that governments have a natural tendency to excessive growth, attempts to
impose some form of tax coordination that limits the effects of cross-border
shopping will not be acceptable. By contrast, if it is considered that governments
attempt to maximize social welfare, tax coordination among jurisdictions will be
desirable. In this regard, the strategy of imposing a minimum tax rate, recommended
by Kanbur and Keen (1993), has been adopted by the European Union for indirect
taxation since the introduction of the Single Market, and is applied in some federal
countries, such as Spain, in order to coordinate the exercise of the tax competencies
of regional governments.
Acknowledgment Julio López-Laborda and Fernando Rodrigo gratefully acknowledge the financial
support from the Ministry of Science and Technology, Project ECO2009-10003.
References
Alamar, B., Mahmoud, L., & Glantz, S. A. (2003). Cigarette smuggling in California: Fact and fiction.
San Francisco: Center for Tobacco Control Research and Education, Tobacco Control Policy Making,
University of California (http://repositories.cdlib.org/ctcre/tcpmus/Smuggling2003).
Alm, J., & Melnik, M. (2005). Sales taxes and the decision to purchase online. Public Finance Review, 33
(2), 184–212.
Asplund, M., Friberg, R., & Wilander, F. (2007). Demand and distance: Evidence on cross–border
shopping. Journal of Public Economics, 91(1–2), 141–157.
Ballard, C. L., & Lee, J. (2007). Internet purchases, cross-border shopping, and sales tax. National Tax
Journal, 60(4), 711–725.
Banfi, S., Filippini, M., & Hunt, L. C. (2005). Fuel tourism in border regions: The case of Switzerland.
Energy Economics, 27(5), 689–707.
Beard, T. R., Gant, P. A., & Saba, R. P. (1997). Border-crossing sales, tax avoidance, and state tax policies:
An application to alcohol. Southern Economic Journal, 64(1), 293–306.
Boisvert, M., & Thirsk, W. (1994). Border taxes, cross–border shopping, and the differential incidence of
the GST. Canadian Tax Journal, 42(5), 1276–1293.
Chiou, L., & Muehlegger, E. J. (2008). Crossing the line: The ef fect of cross border cigarette sales on state excise
tax revenues. John F. Kennedy School of Government Research Working Paper Series RWP08-012
Christiansen, V. (1994). Cross-border shopping and the optimum commodity tax in a competitive and a
monopoly market. Scandinavian Journal of Economics, 96(3), 329–341.
Christiansen, V. (2003). Cross-border shopping and tax structure. EPRU Working Paper Series, 2003-04,
Economic Policy Research Unit (EPRU), University of Copenhagen, Department of Economics.
Coats, R. M. (1995). A note on estimating cross-border effects of state cigarette taxes. National Tax
Journal, 48(4), 573–584.
Crawford, I., & Tanner, S. (1995). Bringing it all back home: Alcohol taxation and cross-border shopping.
Fiscal Studies, 16(2), 94–114.
Crawford, I., Smith, Z., & Tanner, S. (1999). Alcohol taxes, tax revenues and the single European market.
Fiscal Studies, 20(3), 287–304.
Crawford, I., M. Keen and S. Smith (2008). Value-Added Tax and Excises. In Report of a Commission on
Reforming the Tax System for the 21
st
Century, Chaired by Sir James Mirrlees. London: The Institute
for Fiscal Studies.
Di Matteo, L., & Di Matteo, R. (1996). An analysis of Canadian cross-border travel. Annals of Tourism
Research, 23(1), 103–122.
Doyle, J. J., Jr., & Samphantharak, K. (2008). $2.00 Gas! studying the effects of a gas tax moratorium.
Journal of Public Economics, 92(3–4), 869–884.
Ferris, J. S. (2000). The determinants of cross-border shopping: Implications for tax revenues and
institutional change. National Tax Journal, 53(4, Part 1), 801–824.
146 A. Leal et al.
Fisher, R. C. (1980). Local sales taxes: Tax rates differentials, sales loss, and revenue estimation. Public
Finance Quarterly, 8(2), 171–188.
Fisher, R. (2007). State and local public finance (3rd ed.). Mason: Thompson South-Western.
Fitzgerald, J. D. (1992). The distortionary effects of taxes on trade in border areas: The Republic of
Ireland-UK border. In G. Winckler (Ed.), Tax harmonization and financial liberalization in Europe
(pp. 39–56). New York: St. Martin’s Press.
Fleenor, P. (1998). How excise tax differentials affect interstate smuggling and cross-border sales of
cigarettes in the United States. Background Paper, 26, Tax Foundation.
Fleenor, P. (1999). How Excise Tax Differentials Affect Cross-Border Sales of Beer in the United States.
Background Paper, 31, Tax Foundation.
Fox, W. F. (1986). Tax structure and the location of economic activity along state borders. National Tax
Journal, 39(4), 387–401.
Garrett, T. A., & Marsh, T. L. (2002). The revenue impacts of cross-border lottery shopping in the
presence of spatial autocorrelation. Regional Science and Urban Economics, 32(4), 501–519.
Goolsbee, A. (2000). In a world without borders: The impact of taxes on internet commerce. Quarterly
Journal of Economics, 115(2), 561–576.
Goolsbee, A. M., Lovenheim, M. F. & Slemrod, J. (2007). Playing with fire: Cigarettes, taxes and
competition from the internet. Policy Paper, 07-002, Stanford Institute for Economic Policy Research.
Gordon, R. H. (1983). An optimal taxation approach to fiscal federalism. Quarterly Journal of Economics,
98(4), 567–586.
Gordon, R. H., & Nielsen, S. B. (2001). Tax evasion in an open economy: Value-added vs. income
taxation. Journal of Public Economics, 66(2), 173–197.
Hamovitch, W. (1966). Effects of Increases in Sales Tax Rates on Taxable Sales in New York City. In
Financing Government in New York City, Graduate School of Public Administration of New York
University Report to the Temporary Commission on City Finances (pp. 619–634). New York: New
York University.
Haufler, A. (1996). Tax coordination with different preferences for public goods: Conflict or harmony of
interest? International Tax and Public Finance, 3(1), 5–28.
Hvidt, M., & Nielsen, S. B. (2001). Non-cooperative vs. Minimum-rate commodity taxation. German
Economic Review, 2(4), 315–326.
Kanbur, R., & Keen, M. (1993). Jeux Sans Frontières: Tax competition and tax coordination when
countries differ in size. American Economic Review, 83(4), 877–893.
Keen, M. (2002). Some international issues in commodity taxation. Swedish Economic Policy Review, 9
(1), 9–39.
Leal, A., López-Laborda, J., & Rodrigo, F. (2009). Prices, taxes and automotive fuel cross-border
shopping. Energy Economics, 31(2), 225–234.
Levin, H. M. (1966). An Analysis of the Economic Effects of the New York City Sales Tax. In Financing
Government in New York City, Graduate School of Public Administration of New York University Report
to the Temporary Commission on City Finances (pp. 635–691). New York: New York University.
Lockwood, B. (1993). Commodity tax competition under destination and origin principles. Journal of
Public Economics, 52(2), 141–162.
Lovenheim, M. F. (2008). How far to the border?: The extent and impact of cross-border casual cigarette
smuggling. National Tax Journal, 61(1), 7–33.
Maliet, L. D. (1955). A Comparative Study of the Illinois Retailers’ Occupation Tax and the Iowa Retail
Sales and Use Taxes. Unpublished Doctoral Dissertation, University of Illinois, Urbana.
Manuszak, M. D., & Moul, C. C. (2008). How far for a buck? Tax differences and the location of retail
gasoline activity in Southeast Chicagoland, mimeo (http://artsci.wustl.edu/∼moul/pdf_drafts/chicago-
gas-tax.pdf).
Mcallister, H. E. (1961). The border tax problem in Washington. National Tax Journal, 14, 361–374.
Mikesell, J. L. (1970). Central cities and sales tax rate differentials: The border city problem. National Tax
Journal, 23(2), 206–213.
Mikesell, J. L. (1971). Sales taxation and the border county problem. Quarterly Review of Economics and
Business, 11,23–29.
Mikesell, J. L., & Zorn, C. K. (1986). Impact of the sales tax rate on its base: Evidence from a small town.
Public Finance Quarterly, 14(3), 329–338.
Mintz, J., & Tulkens, H. (1986). Commodity tax competition between member states of a federation.
Journal of Public Economics, 29(2), 133–172.
Nielsen, S. B. (2001). A simple model of commodity taxation and cross-border shopping. Scandinavian
Journal of Economics, 103(4), 599–623.
Cross-Border Shopping 147
Nielsen, S. B. (2002). Cross-border shopping from small to large countries. Economic Letters, 77(3), 309–
313.
Ohsawa, Y. (1999). Cross-border shopping and commodity tax competition among governments. Regional
Science and Urban Economics, 29(1), 33–51.
Rietveld, P., Bruinsma, F. R., & Van Vuuren, D. J. (2001). Spatial graduation of fuel taxes; consequences
for cross-border and domestic fuelling. Transportation Research Part A: Policy and Practice, 35(5),
433–457.
Saba, R. P., Beard, T. R., Ekelund, R. B., Jr., & Ressler, R. W. (1995). The demand for cigarette
smuggling. Economic Inquiry, 33(2), 189–202.
Scharf, K. A. (1999). Scale economies in cross-border shopping and commodity taxation. International
Tax and Public Finance, 6(1), 89–99.
Skidmore, M., & Tosun, M. S. (2008). Do new lottery games stimulate retail activity? Evidence from West
Virginia Counties. Journal of Regional Analysis and Policy, 38(2), 45–55.
Smith, R. (1976). The legal and illegal markets for taxed goods: Pure theory and an application to state
government taxation of distilled spirits. Journal of Law and Economics, 19, 393–429.
Thursby, M., Jensen, R., & Thursby, J. (1991). Smuggling, camouflaging, and market structure. Quarterly
Journal of Economics, 106(3), 789–814.
Tosun, M. S., & Skidmore, M. (2004). Interstate competition and lottery revenues. National Tax Journal,
57(2), 163–178.
Tosun, M. S., & Skidmore, M. (2007). Cross-border shopping and the sales tax: An examination of food
purchases in West Virginia. The B.E. Journal of Economic Analysis & Policy, 7: Iss. 1 (Topics),
Article 63 (>http://www.bepress.com/bejeap/ratingsystem.html).
Trandel, G. A. (1994). Interstate commodity tax differentials and the distribution of residents. Journal of
Public Economics, 53(3), 435–457.
Walsh, M. J., & Jones, J. D. (1988). More evidence on the ‘border tax’ effect: The case of West Virginia,
1979–84. National Tax Journal, 41(2), 261–265.
Wang, Y. (1999). Commodity taxes under fiscal competition: Stackelberg equilibrium and optimality.
American Economic Review, 89(4), 974–981.
Warner, K. E. (1982). Cigarette excise taxation and interstate smuggling: An assessment of recent activity.
National Tax Journal, 35(4), 483–490.
Wertz, K. L. (1971). Cigarette taxation by the American States. National Tax Journal, 24(4), 487–492.
148 A. Leal et al.
