306A.Alesina,R.Wacziarg/Journal of Public Economics69(1998)305–321 can share the costs of partially or completely non-rival public goods over larger populations,per capita expenditure on the goods is lower.
2.To the extent that market size influences productivity,large countries can
‘afford’to be clod,while small countries face stronger incentives to remain open;converly,as trade liberalizes,regional and cultural minorities can ‘afford’to split becau political borders do not identify the size of the market;
therefore,smaller countries can enjoy the benefits of cultural homogeneity without suffering the costs associated with small markets(Alesina et al.
(1997)).This hypothesis points toward a negative relationship between country size and the degree of trade openness.
This paper provides empirical evidence consistent with the two ideas.Wefirst show that government consumption,as a share of GDP,is smaller in larger countries.We next confirm the obrvation that small countries tend to be more open to international trade.
The two facts,taken together,may account for the obrvation that open countries have larger governments.In a recent and widely cited paper,Rodrik (1996)suggests a different link between openness and government size.He argues that open countries are more subject to external shocks,and therefore need a larger public ctor to provide a stabilizing role.The prent paper implies a different but not mutually exclusive explanation for the positive empirical relationship between openness and government size.Specifically,we argue that this link is mediated by country size.Hence,we cast some doubts on the direct link between openness and the share of government consumption.On the other hand,wefind some evidence of a direct relationship between openness and the size of government transfers,a result which is in the spirit of Rodrik’s hypothesis concerning the stabilizing role of governments in open economies.
This paper is organized as follows.Section2discuss the argument linking country size,openness and government share and prents some simple statistics on this point.Section3specifies and estimates a more complete t of equations for the determination of government size and trade openness.Section4discuss the evidence concerning the direct effect of openness on government size.The last ction concludes.
2.Size,openness and public goods
2.1.Country size and trade openness
In veral models with increasing returns to the scale of production,market size influences the level of economic activity.We can go back as far as Adam Smith, who argued that the size of the market impos a constraint on the division of labor.Therefore,small countries that are clod to trade must experience a lower
A.Alesina,R.Wacziarg/Journal of Public Economics69(1998)305–321307 overall level of productivity.More recently,Murphy et al.(1989)propo a model of industrial development in which market size determines the extent to which firms can benefit from positive spillovers from each other.In this model,low income countries may need a‘Big Push’in order to move from a‘bad’equilibrium characterized by traditional,constant returns technologies to a‘good’equilibrium with modern,increasing returns industries.Ades and Glaer(1994); Wacziarg(1997);Alesina et al.(1997)provide empirical evidence consistent with the ideas:large countries experience smaller dynamic gains from trade than smaller countries.
In a world without international trade,political boundaries identify markets and countries face economic incentives to be large.On the contrary,the more a country can trade with the rest of the wor
ld,the less one can identify its political borders with the boundaries of its market.This obrvation has two implications: as the world trade regime becomes more and more open,various ethnic groups and regions willfind it feasible to break away from their original countries;more generally,countries willfind it less costly to split.Converly,as the world becomes more and more populated by small countries,a liberal trade regime will find more and more supporters,precily becau small countries need trade to be economically viable.In other words,small countries face incentives to adopt open trade policies,precily becau they cannot benefit from access to larger markets unless they are open to trade.Thus,small countries can be expected to be more open to trade.
Alesina et al.(1997)document the effect of trade openness on country , on cessions and mergers.They start from some well known facts.After the Second World War,in a period of rapidly increasing trade liberalization,the number of countries incread from74in1946to192in1995.In1995,87 countries had less than5million inhabitants,58less than2.5million and35less than500000.More than half of the world’s countries are smaller(in population)
1
than the state of Massachutts.Many factors have contributed to this develop-ment,particularly the decolonization of Africa and the collap of the Soviet Union.However,the trade regime is also important.For instance,veral new small countries that emerged in the former Eastern bloc may not have chon
2
independence in a world of heavy trade restrictions.
Other arguments also point to the fact that smaller countries should trade more, with causation running from country size to obrved trade volumes directly, rather than through trade policy.A simple way to illustrate this type of argument is to undertake a simple thought experiment:Consider a large country living in
1In1990the State of Massachutts had a population of6016425.Ninety eight countries have smaller populations.
2Note that veral new countries in Eastern Europe and the former Soviet Union are quite small.For instance,Latvia has1.7million inhabitants,Turkmenistan4million,Moldova4.5million and the Kyrgyz Republic4.8million.
308A.Alesina,R.Wacziarg/Journal of Public Economics69(1998)305–321 autarky;if this country breaks up into smaller,free-trading units,each new unit will suddenly exhibit positive international trade.
2.2.Country size and the size of government
To the extent that there arefixed costs and economies of scale linked to partial or complete non-rivalry in the supply of public goods,smaller countries may have a larger share of government in GDP.For instance,there arefixed costs in establishing a t of institutions,a legal,monetary,andfiscal system.At least up to a point,when congestion effects become relevant,the costs of certain public goods grow less than proportionally to the size of the population(parks,libraries,roads,
3
telecom infrastructures).
To the extent that public goods are of a non-rival nature,increasing returns stem from the fact that,while the required level of provision is independent of population size(or grows less than proportionately to it in the ca of partial non-rivalry),the cost of public goods can be spread over a la
rger pool of taxpayers in larger countries.The following simple example illustrates this point:Consider a country compod of N identical individuals with constant elasticity of substitution utility functions.The social planner maximizes the utility of a reprentative individual:
a a1/a
U5(C1G)(a#1)(1) where C is private consumption and G is a non-rival public good.If the size of the population is N,Y is the exogenously given level of individual income and taxes are lump-sum,then the individual budget constraint will be:
G
]
C5Y2(2) N
The non-rival nature of the public good implies that every agent derives utility from consuming its aggregate supply G.However,each individual only pays a fraction1/N of the total cost.Thefirst order condition obtained from maximizing equation(1)subject to equation(2)leads to the following optimal supply of G:
YN
]]]
G5(3) a
]
N11
a21
This implies that the ratio of government spending to aggregate GDP,which is our variable of interest,is the following:
3For some of the public goods,population density is also a critical factor(we control for this variable in our empirical analysis).
A .Alesina ,R .Wacziarg /Journal of Public Economics 69(1998)305–321309
G 1]]]]5(4)a YN ]a 21N 11
and:
1]a 21≠(G /YN )a N ]]]]]]]]]52S D (5)a 2≠N a 21]S a 21D N 11This expression is negative whenever a ,0.The less substitutable C and G (a →2`),the more we approach the ca of a Leontief utility function,and the greater the effect of population on the government spending to GDP ratio.On the contrary,in the ca of a unit elasticity of substitution (a 50),the utility function approaches a Cobb-Douglas and the effect of country size becomes zero.As the elasticity of substitution keeps increasing (a 51corresponds to linear utility),the effect of population becomes positive.
The intuition here is that an increa in country size has two effects:it reduces the per capita cost of public goods for a given level of provision,allowing more private consumption,which corresponds roughly to an income effect,and it rais the optimal level of provision (this is akin to a substitution effect).The more substitutable private consumption and public goods,the more agents will be willing to increa their level of consumption of the public good as a result of a decrea in its per capita cost.In this ca,the substitution effect dominates and country size will actually be positively related to the ratio of government spending to GDP.The empirical test for whether increasing returns to public goods provision lead to a smaller government to GDP ratio is esntially a test of whether the right-hand side of equation (5)is negative.
In summary,we will test for an inver relationship between the size of a country and the share of government consumption and investment,that is,we will bring equation 5to the data.Note that this argument is most relevant for government consumption of goods and rvices,while transfers should not be included in the definition of government spending for which increasing returns should apply.
2.3.Some basic statistics
Table 1describes all the variables ud in this paper.We measure openness mainly as the share of imports and exports over GDP.Our focus is on actual trade integration,which captures access to wider markets and also includes gravity effects,since we are interested in the fraction of the economy which actually ‘interacts’with the rest of the world.However,we also examined the relationship between trade policy and country size,using tariff rates as well as other available indicators of outward orientation.
