ChapterPDF Available

Why Do Weak States Prefer Prohibition to Taxation?


Abstract and Figures

Why do weak states prefer prohibition to taxation? Desierto and Nye show that keeping an undesirable good illegal is more efficient than legalizing and taxing it, even if producers of the prohibited goods pay out large bribes to prohibition enforcers. If the bribes are recognized as revenues to the enforcers, this additional benefit keeps welfare losses small. This chapter further supports this finding with preliminary empirical evidence and graphical analyses of the likely net welfare losses from prohibition and taxation. It provides a positive rationale for the preference for prohibition in states prone to corruption and imperfect enforcement.
Content may be subject to copyright.
Why Do Weak States Prefer
Prohibition to Taxation?
21 September 2010
Why do weak states prefer prohibition to taxation? Despite
their relative inability to enforce prohibitions on many illegal ac-
tivities such as gambling or selling narcotics, most states choose to
prohibit rather than tax these goods and services. Moreover, na-
tions marked by widespread corruption or low state capacity are
more likely than e¤ective states to prefer prohibition. Desierto
and Nye [2010] show that keeping an undesirable good illegal is
more e¢ cient than legalizing and taxing it, even if producers of
prohibited goods pay out large bribes to state enforcers. If the
bribes are internalized as revenues to enforcers, this additional
bene…t actually shrinks net welfare loss. This paper combines
this …nding with empirical work on cross-national support for
prohibition. Graphical analyses illustrate the net welfare losses
of prohibition vs. taxation and help establish this new positive
rationale for the persistence of prohibition worldwide.
1 Introduction
Proponents of eliminating the ban on illegal narcotics have long noted
that prohibition is often ine¤ective and counterproductive. Recent work
in the economics literature indicates that prohibiting “undesirable”goods
such as drugs is ine¢ cient (cf. Miron [2004], [2008] and Becker, Murphy,
and Grossman (BMG) [2006]). BMG is especially notable for discussing
D. Desierto: University of the Philippines, School of Economics, Diliman, Q.C.
1101, email:; J. Nye: Economics Department, George
Mason University, Carow Hall MSN 1D3, 4400 University Drive, Fairfax VA 22030,
email: Research assistance provided by Karen Annette D. Lazaro.
the problem of controlling the spread of an illegal good when enforce-
ment is imperfect and avoidance costs are factored in. BMG show that
prohibition can only be justi…able if the good is really undesirable, that
is, if the marginal social value of consuming the good is very low or, in
the case of inelastic demand, su¢ ciently negative. This is essentially
because illegal good producers waste resources in order to avoid being
detected, captured and/or penalized. They can, for instance, bribe pro-
hibition enforcers - from police o¢ cers to courts - in order to continue
supplying the market. Such avoidance costs, as BMG imply, are a dead-
weight loss to society. On the other hand, legalization and taxation of
the good are a more e¢ cient way of curbing consumption, since taxes
paid for the good are eventually plowed back to society. In other words,
avoidance costs are a leakage, while taxes are not.
Desierto and Nye [2010] show, however, that avoidance costs are not
necessarily a loss, if bribes are treated as additional income or revenue
to enforcers. And, in contrast, taxes can be a loss since the government
might use tax revenues ine¢ ciently and might even illegally appropriate
some of it. In this latter case, losses might be limited if the appropriated
amount is also treated as additional revenue to tax enforcers, but cor-
ruption by tax enforcers is less likely than the corruption committed by
prohibition enforcers. Bribery between illegal good producers and pro-
hibition enforcers is likely to be more sustainable because both are com-
plicit in the corruption activity, while legal good producers/taxpayers
have little incentive to allow tax enforcers to illegally appropriate the
Desierto and Nye thus show that in a second-best world in which cor-
ruption is present, prohibition is likely to be more e¢ cient than taxation.
This paper provides further support for this result.
None of the existing literature has dealt with the positive question
of why undesirable goods are so often prohibited if enforcement is so
lax. In contrast, our approach suggests why prohibiting undesirable
goods like illegal narcotics or gambling is especially common in places
with weak institutions and a reputation for corruption. In Section 2, we
present empirical evidence indicating that not only is prohibition a more
commonly employed method of curbing consumption of drugs, prostitu-
tion, and gambling than legalization and taxation, but that prohibition
is more prevalent in countries with greater corruption. Section 3 then
graphically analyzes equlibria in the BMG and Desierto and Nye mod-
els, and then compares the likely losses from prohibition vs. taxation.
Section 4 concludes and thus provides a positive rationale for prohibition
and an e¢ ciency motivation for prevalence of prohibition in weak states.
2 Prohibition and Corruption
Despite the supposed bene…ts of legalization with taxation, prohibition
is widespread. Among a sample of 101 countries that we have obtained,
100 prohibit drugs, 66 prohibit prostitution, and 33 prohibit gambling.1
Of course, this fact alone need not be inconsistent with BMG’s results -
it might just be that the social marginal values of consumption for these
goods are very low. Note, however, that some countries that are roughly
comparable in terms of levels of development, culture, and/or geograph-
ical location, and thus have arguably similar social marginal values of
consumption of undesirable goods like drugs, prostitution and gambling,
can still have di¤erent approaches to curbing production/consumption
of these goods. For instance, drugs are illegal in the US, Canada and
most of Europe, but are legal in the Netherlands; prostitution is legal in
France, but not in Switzerland; gambling is prohibited in Thailand, but
not in the Philippines.
In addition, some illegal goods, e.g. gambling, pirated products,
might pose little negative consumption externalities and might even have
high consumption value and/or produce positive externalities, and yet
they are still persistently prohibited. It is not clear, therefore, that
the BMG result is the only explanation for why prohibition seems to
be a sustainable equilibrium. As Desierto and Nye posit, bribes and
rent-seeking by corrupt enforcers are easier to extract when markets are
Indeed, prohibition seems to be more common in countries that are
more prone to corruption and that have weaker state capacities than
those which merely tax. We obtain the 2009 Corruption Perceptions
Index (CPI) from Transparency International for each country in the
sample and compute the average CPI for countries that prohibit, and
for those that legalize and tax, drugs, prostitution and gambling.2The
following tables summarize the results and clearly suggest that corrup-
tion is higher (i.e. the average corruption score is lower) among countries
that prohibit, than among those that legalize and tax:
1See Appendix for details.
2CPI scores re‡ect "political stability, long-established con‡ict-of-interest regula-
tions and solid, functioning public institutions".
Prohibition Taxation
Number 100 1
Ave. Corruption Score 3.979 8.9
(Range: 9.4 to 1.1)
Table 1
Prohibition Taxation
Number 66 35
Ave. Corruption Score 3.53 4.96
(Range: 9.4 to 1.1)
Table 2
Prohibition Taxation
Number 33 68
Ave. Corruption Score 3.31 4.26
(Range: 9.4 to 1.1)
Table 3
3 Graphical Analysis
In BMG and Desierto and Nye, illegal producers can o¤er bribes to pro-
hibition enforcers as part of total avoidance costs AC that they incur.3
The illegal producer chooses the level of AC that minimizes its expected
cost, given the level of prohibition/enforcement e¤ort Ethat the gov-
ernment undertakes. Meanwhile, given the amount AC that the illegal
producer spends, the government chooses its level of Ethat maximizes
social welfare W: The main di¤erence between BMG and Desierto and
Nye, however, is that while the former assumes that the government
can/does only maximize consumers’and producers’welfare, the latter
acknowledges that the government might also maximize the bene…ts to
enforcers (being, ultimately or indirectly, also consumers and/or pro-
ducers of the good). That is, unlike BMG, Desierto and Nye assume
that the goverment includes the welfare of enforcers, and treats bribes as
additional revenues: W=V+R+B; where Vis social value of consump-
tion (net of all other consumption externalities), Rare the producers’
net revenues, and Bare (net) bribe revenues of enforcers.
Solving simultaneously for AC and E, equilibrium is achieved at the
point where: V q Bq =M R; where V q is the marginal social value of
consumption, Bq is marginal bribe revenue to enforcers, and M R is mar-
ginal revenue of producers. In contrast, equilibrium in BMG is captured
by: V q =M R: Thus, ’internalizing’bribes tends to achieve higher AC
and E, precisely because the bribe bene…ts provides additional incentive
to prohibition enforcers, which then requires illegal producers to spend
more to counteract the increased enforcement e¤orts. The optimal level
of consumption is thus lower, making prohibition more e¤ective in cur-
tailing consumption, and more e¢ cient since it need not incur additional
losses in decreasing consumption further - the additional avoidance costs
in the form of bribes are not ’wasted’but goes to enforcers as bribe rev-
The following graphs illustrate this result. We …rst consider the
corner-solution case, in which either of two extremes is socially optimal
- completely freeing the market (i.e. legalization), or full, all-out en-
forcement against the good which drives consumption to zero. We show
that internalizing bribes makes the latter more likely to be optimal than
the former. The next case depicts an interior solution, where some im-
perfect level of enforcement is optimal, allowing some positive level of
consumption. Here it is shown that internalizing bribes actually de-
3Desierto and Nye, however, make the bribe amount explicit, by assuming that
a …xed fraction of total AC are in the form of bribes. This fraction captures the
overall extent of corruption in the environment, as it is the permissible level at which
enforcers can extract a bribe without being detected by the government.
creases optimal consumption e¢ ciently. We then compare this outcome
to taxation as an alternative method of restricting consumption, and
show that taxation is likely to be less e¢ cient than prohibition with (in-
ternalized) bribery. The result holds even when tax collectors/enforcers
are also corrupt (like prohibition enforcers), and the corruption is inter-
nalized. This is essentially because the internalization of corruption is
limited when the good is legal than when it is prohibited - it is more
di¢ cult for tax enforcers to (illegally) appropriate taxes paid by legal
producers than for prohibition enforcers to extract bribes from produc-
ers who have to keep quiet to stay underground.
3.1 Full Enforcement vs. Legalization
The graph depicted in Figure 1 reproduces BMGs Figure 2:Quantity
Qu; although it satis…es V q =MR, is not socially optimal as it violates
the 2nd-order condition. (To the right of Qu,V q falls slower than MR,
so it makes sense to keep increasing output. To the left of Qu,MR rises
faster than V q, so it makes sense to keep decreasing output.) Thus,
the only possible optimal consumption levels are either zero (i.e. full
enforcement) or the free-market level Qf (i.e. zero enforcement) at which
demand Dis equal to the marginal cost Cof producing the good. This
depends on the relative gains depicted by the triangles to the left and
to the right of Qu - if triangle bis larger than triangle a, then it is
more socially optimal to legalize the good, at which case Qf is achieved.
Otherwise, it is better to fully prohibit the good to curtail consumption
to zero.
Qu Qf
Figure 1: BMG equilibrium (corner solution)
Figure 2 depicts equilibrium in the Desierto-Nye model, that is, when
bribes to enforcers are internalized. At quantity Qb,V q Bq =MR,
but this also violates the 2nd-order condition. Thus, as in BMG, either
the zero or free-market consumption level is optimal, depending on the
relative social gains as captured by the triangles between (V q Bq)and
MR to the right and to the left of Qb. Figure 3 labels these areas cand
d: In this example, dis smaller than c; which makes full enforcement
(zero consumption) socially optimal. More generally, however, the in-
ternalization of bribes always makes area c>aand d < b. That is,
the bene…ts of legalization shrinks vis-à-vis the bene…ts of (full) enforce-
ment, thus making free market consumption less likely to be optimal.
This is because legalization foregoes the opportunity of enforcers to earn
additional revenue by extracting bribes when the good is prohibited.4
4With Bq > 0, it is always the case that (V q Bq)is less than, or below, V q:
In Figures 2 and 3, the (V q Bq)line has roughly the same slope as the V q line,
but this is not necessarily the case. While the (V q Bq)line cannot be ‡atter than
V q (i.e. they cannot intersect), it can be steeper. This can happen if Bq is more
sensitive to output changes than V q - that is, if deriving bribe bene…ts is more costly
Qu Qf
Figure 2: Desierto-Nye equilibrium (corner solution)
than enjoying ’legitimate’consumption. In this case, the result is stronger, since
area dwould be smaller, and area clarger.
Qu Qf
Figure 3: With bribery, legalization is less likely to be optimal.
3.2 Imperfect Enforcement
When the 2nd-order condition is satis…ed at V q =MR, a non-zero but
restricted level of consumption is socially optimal, and some imperfect
level of enforcement is justi…able. In Figure 4, the V q line is now
steeper than MR. At low levels of quantity, each extra good is valued
more by society (as consumption good) than by producers (as source
of revenue). After Qu, each additional good becomes more costly to
society than to producers.5Hence, the socially optimal consumption
level is achieved at Qu;(which represents some imperfect enforcement
level), with corresponding price P.
5Or, to state it analogously with the previous subsection: to the right of Qu,V q is
falling faster than MR, so it is not socially optimal to increase quantity further. To
the left of Qu,V q is rising faster than M R, so it is not optimal to decrease quantity.
Qu Qf
Figure 4: BMG equilibrium (interior solution)
Note that producers are perfectly competitive and enjoy zero pro…ts.
Their total revenue, given by area f+ein Figure 5, is used to cover
the total cost of manufacturing Qu; which is captured by area e, and
avoidance costs’, as depicted by area f. There is a loss of consumer
surplus, captured by area g, since not all a¤ordable demand is served.
Qu Qf
Figure 5
Figure 6 depicts the interior solution in the Desierto-Nye model. It
can be seen that when bribery is taken into account, the optimal con-
sumption level given by (V q Bq) = MR is lower at Qb < Qu. This is
because enforcers have an incentive to increase e¤ort. The higher e¤ort
is re‡ected in higher price P0- that is, illegal producers spend more on
avoidance per unit in order to counter the increased enforcement. Note
that since some avoidance costs are spent on bribes, and the bribes are
internalized, this fraction of avoidance costs is not really wasted, but
just goes to enforcers as their revenues.
Qu Qf
Figure 6: Desierto-Nye equilibrium (interior solution)
As Figure 7 shows, restricting consumption to Qb entails enforcers
exerting a high enough e¤ort (re‡ected in P0) and producers spending
area i+kto be able to sell Qb amount to consumers without getting
caught. (That is, of course, after manufacturing Qb and incurring cost
area h:)Out of i+k, area kgoes to enforcers as net bribes (while i
might be spent on litigation, relocation, and other ways to avoid being
caught and punished). The gross bribe revenue is given by area j+k
as though the producer spends an equivalent of this to pay for the bribe
and thus incurs additional manufacturing cost jplus premium k. In this
manner, the extra production of Qu Qb is ’disguised insofar as it is
used only to pay for the bribe. Or to put it di¤erently, the enforcer
receives bribes of (Qu Qb)amount of drugs and can sell it to other
producers at a premium and get pro…t equal to k.
Qu Qf
Figure 7: Prohibition is more e¢ cient when bribery is internalized.
Thus, while consumption is curtailed at Qb, all the production re-
sources are equivalent to producing Qu. This, though, is not a waste
it just goes to enforcers. Enforcement is more e¤ective and e¢ cient
when bribery is taken into account since curtailing consumption to Qb
does not incur any additional losses greater than the area g. In other
words, if Qb were the target consumption level (in both the BMG and
Desierto-Nye model), not internalizing bribery (BMG result) would in-
cur losses the total of k+g, while internalizing it (Desierto-Nye result)
would only incur loss g.6
3.3 Imperfect Enforcement vs. Taxation
Suppose we again assume that Qb is the target optimal consumption
level. Then a high-enough tax rate tcan also achieve this, as illustrated
in Figure 8. The total revenue is given by area i+h, where hcovers
manufacturing costs of Qb, while iare paid as taxes to the government.
The potential loss of consumer surplus is k+g, which is bigger than the
6Recall that in Figure 4, area gis lost when the optimal consumption level is
Qu:If optimal consumption were less than this, the loss would be greater than g.
loss gincurred by prohibition with internalized bribery (in Figure 7).
However, if tax revenues are used e¢ ciently by the government, this loss
may be o¤set or minimized. Government investments and spending on
public goods and services might be ’pro…table’such that an equivalent
of ior more is transferred back to society. (In the example in Figure
8, it turns out that a pure, ’one-for-one’, transfer will still incur losses
since area iis smaller than area k+g. To eliminate all losses, the tax
revenues ihave to be used pro…tably enough such that they have returns
the equivalent of k+g.)
Qu Qf
C+t, P
Figure 8
The question, then, is how much of area ican be used e¢ ciently to
cover some or all of k+g, or, at least an area equivalent to k, to make
taxation as e¢ cient as prohibition with internalized bribery. The likely
answer is that taxation may not be as e¢ cient. In certain cases, area k
might be bigger than area i, which means that even a pure transfer of
tax revenues might not achieve the same e¢ ciency as prohibition with
bribery. It would then take more than a pure transfer, but this is unlikely
if government projects/investments have low returns.
But suppose we let corrupt tax enforcers apportion tax revenues,
and internalize these illegal appropriations as additional revenues to tax
enforcers. Would the added incentive to corrupt tax enforcers induce
them to exert more tax-collecting e¤orts so as to curb more consumption
without generating further e¢ ciency losses? That is, can we have the
following situation in Figure 9, which is exactly equivalent to the result
obtained by prohibition with internalized bribery? If a total of i+ktax
revenues can be collected, but area kis appropriated by tax enforcers
as ‘net illegal tax revenue’, then the loss is just area gthe same as in
the prohibition with bribery case –and taxation becomes as e¢ cient as
the latter. (It is also as e¤ective, since Qb is still optimal consumption
- additional (Qu Qb)is produced only to cover the equivalent of what
tax enforcers would appropriate from tax revenues.)
Qu Qf
C+t, P
k g
Figure 9
This situation, however, is unlikely. The more likely result is shown
by Figure 10, where only area lcan be illegally appropriated by tax
enforcers, and only area inends up in government co¤ers. That is,
not only is the receipt of the government smaller, but also what tax
enforcers can illegally appropriate. Thus, compared to prohibition with
bribery (which has loss gin Figure 7), taxation with illegal appropriation
incurs a loss of n+m+g.
Qu Qf
Figure 10: Taxation is less e¢ cient than prohibition with internalized
The reason for this is that the price of the good when some taxes are
appropriated drops down uniformly to C+t0, while in the prohibition
case, the e¤ective price of the bribe is, as it were, discriminatory. (Notice
in Figure 7 that price goes up as the bribe quantity goes down.) In
contrast, when tax enforcers’ corrupt activities are internalized, legal
producers have to produce up to Qu to ’fund’illegal appropriation, but
price-discrimination is d cult, if not altogether impossible, because it
is hard to ’disguiseQb Qu. Corruption’quantities, i.e. those beyond
Qb; are priced the same as ’pure consumption’ quantities, i.e. those
until Qb; since all these units are homogeneous to legal producers and/or
consumers. That is, the latter have very little incentive to allow corrupt
enforcers to appropriate taxes, or equivalently, to allow Qb Qu to go
unnoticed’. Thus, this additional supply in circulation induces price to
go down (uniformly) to C+t0:In the prohibition case, however, illegal
producers are complicit with bribe-takers - they do not want to get
caught, afterall - and, hence, each bribe unit in Qb Qu can remain
hidden’and can then be priced discriminately.
Furthermore, tax collectors can rarely appropriate tax revenues only
from the good. If collectors are corrupt, they may appropriate all other
kinds of tax revenue. Or they may end up appropriating revenues from
some goods, but not from others. In this case, the benets of tax en-
forcers from illegally appropriating tax revenues from the good in ques-
tion are di¢ cult to isolate and, hence, Qb Qu are not distinguishable
from Qb: In contrast, bribery revenues from prohibited goods are more
easily identi…ed and isolated which makes Qb Qu distinct from Qb.7
Again, this allows for price-discrimination of the corruption goods in the
prohibition case, but not in the taxation case.
In other words, it is more di¢ cult to internalize corruption when
the good is legal and taxed, than when it is prohibited. This makes
corruption more e¢ cient in illegal, than in legal, environments.
4 Conclusions
Prohibition of undesirable goods can be a rational response especially
if governments are corrupt and enforcement is imperfect. Making a
good illegal provides prohibition enforcers opportunities to extract a
bribe from illegal producers. Since the latter are willing to pay the
bribes, a sustainable equilibrium is achieved whereby enforcers keep their
orts high and illegal producers spend more to stay in business. While
this is seen by BMG as a ’waste’of resources, Desierto and Nye argue
that this is not necessarily the case, if the bribes received by enforcers
are internalized as ’bribe revenues. The result is that, for the same
amount of resources spent by illegal producers so as not to get caught
and/or punished, spending more of it as bribes to enforcers reduces
optimal consumption more than if such spending were apportioned on
other ’external avoidance activities. This is essentially because it is
easier to internalize enforcers’bribe revenues than, say, the additional
income of lawyers hired by illegal producers. Such lawyers cannot easily
ect the amount of illegal goods that end up in the market, whereas
enforcers directly determine it.
The same result holds even when compared to taxation. Tax collec-
tors can indeed determine the amount of tax that goes to the government
and, hence, if they are incentivized by the possibility of illegally appro-
priating some of the tax revenues, they might end up increasing tax en-
forcement e¤orts, which would then decrease production/consumption
7Glaeser and Shleifer (2001) similarly note that is is easier for an enforcer to detect
posession of an illegal good than to verify whether taxes (on a legal good) have been
of the good. However, this is a less e¢ cient way of curbing consumption
because tax collectors cannot appropriate tax revenues as easily as pro-
hibition enforcers can extract bribes from illegal producers. In illegal
markets, bribes and rent-seeking are more easily masked, whereas legal
markets are more transparent. Thus, corrupt prohibition enforcers have
greater incentive to enforce against the ’undesirable’good than corrupt
tax collectors. The latter care less about enforcing the tax on the good
since their corrupt action can be more easily detected or reported and
they need not focus on collecting taxes only on that good as there are
other sources of tax revenues and tax collectors rarely specialize on any
one undesirable good.
[1] Becker, Gary S., Murphy, Kevin M. and Grossman, Michael. 2006.
"The Market for Illegal Goods: The Case of Drugs." Journal of Po-
litical Economy 114:1, 38-60.
[2] Desierto, Desiree A., and John V.C. Nye. 2010. "The Market for
Illegal Goods in the Presence of Corruption." Working Paper, UP
School of Economics and George Mason University.
[3] Glaeser, Edward L., and Andrei Shleifer. 2001. “A Reason for Quan-
tity Regulation.”A.E.R. Papers and Proc. 91 (May): 43135.
[4] Miron, Je¤rey A. 2008. "The Budgetary Implications of Drug Prohi-
bition." Report for the Criminal Justice Policy Foundation and Law
Enforcement Against Prohibition.
[5] Miron, Je¤rey A. 2004. Drug War Crimes: The Consequences of
Prohibition. Oakland, CA: Independent Inst.
CPI 2009 Score
(highest - least corrupt) Prohibition Taxation Prohibition Taxation Prohibition Taxation
New Zealand Developed 9.4  
Denmark Developed 9.3  
Singapore Developed 9.2 + +
Switzerland Developed 9  
Netherlands Developed 8.9  
Canada Developed 8.7 +
Luxembourg Developed 8.2   +
Germany Developed 8  
Ireland Developed 8  
Japan Developed 7.7 
United Kingdom Developed 7.7 +
United States Developed 7.5 +
Qatar Developed 7
France Developed 6.9  
Chile Developing 6.7  
Uruguay Developing 6.7  
Estonia Developing 6.6 +
UAE Developed 6.5 
Israel Developed 6.1 +
Spain Developed 6.1 +
Portugal Developed 5.9 +
Botswana Developing 5.6  
Brunei Developing 5.5 
South Korea Developed 5.5  
Mauritius Developing 5.4  
Bahrain Developing 5.1 
Hungary Developing 5.1  
Jordan Developing 5
Czech Republic Developed 4.9  
South Africa Developing 4.8  
Malaysia Developing 4.5  
Namibia Developing 4.5 +
Cuba Developing 4.4 
Turkey Developing 4.4  
Italy Developed 4.3  
Saudi Arabia Developing 4.3 
Tunisia Developing 4.2  
Georgia Developing 4.1  
Kuwait Developed 4.1 
Ghana Developing 3.9  
Bulgaria Developing 3.8 +
Greece Developed 3.8  
Brazil Developing 3.7  
Peru Developing 3.7  
China Developing 3.6  
Serbia Developing 3.5  
El Salvador Developing 3.4 +
India Developing 3.4 
Thailand Developing 3.4 
Malawi Least Developed 3.3  
Mexico Developing 3.3  
Rwanda Least Developed 3.3  
Albania Developing 3.2  
Liberia Least Developed 3.1  
Sri Lanka Developing 3.1  
Jamaica Developing 3
Senegal Least Developed 3  
Zambia Least Developed 3  
Argentina Developing 2.9 +  
Niger Least Developed 2.9 
Algeria Developing 2.8 
Egypt Developing 2.8  
Indonesia Developing 2.8 
Mali Least Developed 2.8  
Armenia Developing 2.7  
Ethiopia Least Developed 2.7  
Kazhakstan Developing 2.7  
Vietnam Developing 2.7  
Guyana Developing 2.6  
Syria Developing 2.6 
Honduras Developing 2.5  
Lebanon Developing 2.5  
Maldives Developing 2.5 
Mozambique Least Developed 2.5  
NIcaragua Developing 2.5  
Bangladesh Developing 2.4  
Pakistan Developing 2.4 
Philippines Developing 2.4  
Azerbaijan Developing 2.3 
Nepal Developing 2.3 
Ecuador Developing 2.2  
Kenya Developing 2.2  
Russia Developing 2.2  
Ukraine Developing 2.2  
Zimbabwe Least Developed 2.2 
Papua New Guinea Developing 2.1  
Paraguay Developing 2.1  
Yemen Developing 2.1 
Cambodia Developing 2  
Laos Developing 2  
Angola Developing 1.9 
Kyrgyzstan Developing 1.9 +
Venezuela Developing 1.9  
Haiti Least Developed 1.8  
Iran Developing 1.8 
Uzbekistan Developing 1.7 
Chad Least Developed 1.6 
Iraq Developing 1.5  
Myanmar Developing 1.4  
Afghanistan Least Developed 1.3 
Somalia Least Developed 1.1 
+Some form of partial prohibition.
See the following sources for details:
Drugs ,,¤ences/drug_o¤ences¤er/LIBRARY/southam1.htm,
New-Guinea.php¢ ckers.html
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
This paper considers the costs of reducing consumption of a good by making its production illegal and punishing apprehended illegal producers. We use illegal drugs as a prominent example. We show that the more inelastic either demand for or supply of a good is, the greater the increase in social cost from further reducing its production by greater enforcement efforts. So optimal public expenditures on apprehension and conviction of illegal suppliers depend not only on the difference between the social and private values from consumption but also on these elasticities. When demand and supply are not too elastic, it does not pay to enforce any prohibition unless the social value is negative. We also show that a monetary tax could cause a greater reduction in output and increase in price than optimal enforcement against the same good would if it were illegal, even though some producers may go underground to avoid a monetary tax. When enforcement is costly, excise taxes and quantity restrictions are not equivalent.
With corruption, prohibition is better at curtailing consumption than taxation. Prohibition enforcers are incentivized to enforce against illegal producers to extract bribes from them, while the latter willingly pay the bribes to keep supplying the market. In equilibrium, total quantity is lower. In contrast, tax collectors have less incentive to restrict quantity (by imposing higher taxes), even if they were to illegally appropriate tax revenues. Legal taxpayers have less reason to cooperate with corrupt tax collectors, and are more likely to report them. Because corruption is harder to sustain in a legalized market, taxation is less effective.
Drug war crimes: the consequences of prohibition
  • J A Miron
  • JA Miron
The budgetary implications of drug prohibition Report for the criminal justice policy foundation and law enforcement against prohibition
  • Ja Miron
Report for the criminal justice policy foundation and law enforcement against prohibition
  • J A Miron