Turkan Aliyeva Karimova

Capital taxation and globalization

Isabel H. Correia

Abstracts

The decline of capital taxation is associated with efficiency gains. We show that, when
agents are heterogeneous, equity concerns can change the policy recommendation driven by
efficiency. Given the empirical evidence on the roots of heterogeneity inside each country,
either in developing or developed economies, the elimination of capital taxation would lead
always to a decline in inequality and to an increase of welfare of the poorest, in a small open
economy acting unilaterally. On the contrary for a closed economy, or for group of open
economies following the same policy, the opposite can be the result: with the elimination of
capital taxation it can hurts the poorest of each country. Therefore a low degree of capital
openness can support a positive tax on capital.

Introduction

In times of low growth and low investment, policies that can improve aggregate
investment and aggregate output are especially welcome. Given the high level of public debts
in most developed countries, however, the measures under scrutiny should pay for
themselves, and should not require alternative finnancing for the government.Increased
international market integration –also known as globalization-has affected significantly the
design and the scope of fiscal policy. Focusing on factor income taxation, theory suggests that
international factor mobility leads national government, in an attempt to attract mobile
factors, to cut the tax rate on the relatively mobile factors- capital- and increase the tax burden
fallen on the relative immobile factors. Although there is a large number of empirical studies
examining the effect of increased international market integration on national tax policy, the
results of the relevant literature appear to be rather inconclusive. This fact is even more
puzzling when we take into account that the tax on capital income, when compared either
with the tax on labor income or the value-added tax, is much more ineffi cient. This is a well
known and robust result in the literature. Taxing capital imposes a negative incentive on
saving (that is, an intertemporal distortion since it taxes more heavily future than current
consumption). This
characteristic is worsened with the double, or sometimes triple, capital taxation that
Turkan Aliyeva Karimova
characterizes most fiscal codes. Social welfare is usually pointed out as the reason for the
relatively high taxes on capital income, due to the undesirable effects on equity that a decline
of those taxes could deliver, namely when compensated by an increase of labor taxation.
Therefore, the existing situation can be seen as an implicit and partial coordination system,
since different countries have different levels for capital taxes. This arrangement is supported
by the argument that lower taxes would have a positive impact on effi ciency, at the cost of
penalizing the poorest of each economy. This equity loss is a cost that most countries would
not want to pay.Regarding capital taxation, a branch of the empirical literature concludes that
higher international market integration is associated with higher capital taxation.
This article uses a general equilibrium framework to show that the effect on equity
caused by the elimination of the tax on capital could theoretically depend mainly on the joint
distribution of characteristics that determine the society's heterogeneity.
However, using the empirical evidence on crosssectional distribution, we show that this result
is mainly driven by the degree of effective international mobility of capital. In the case of a
small open economy with perfect capital mobility, which decides unilaterally to change
policy, we show that inequality is reduced. The intuition for this result is simple: when capital
taxation is eliminated, the net return on capital declines in the rst period, and net wages in the
new steady state are always higher, since the effect of capital in ows on the marginal
productivity of labor dominates the higher tax on labor. Since these wages are iscounted using
the international real interest rate, which is exogenous to policy, the total present value of
labor income increases.
Therefore, when agents differ in both wealth or labor eficiency, but wealth is more
unevenly distributed than earnings, the poorest agents are always better after the elimination
of capital taxation. This result is in clear contradiction with the one in Garcia-Milla, Marcet
and Ventura (2010) or Domeij and Heathcote (2004) both of which use a closed economy
model. It is also in contradiction with the more popular argument based on a partial
equilibrium reasoning: the reduction of the tax on capital and the increase of the tax on labor
income increases the return on capital and decreases the return on labor, and therefore benefits
the upper income agents and harms the lower income agents. Therefore we try to understand
why the degree of capital globalization is a major determinant of how the elimination of
capital taxation affects inequality. To understand the contradiction, we use a closed
economy model, similar to the one we had for the open economy, and we repeat the exercise.
This allows us to clarify the apparent contradiction between our results and theirs. We show
that the difference arises because, in the small open economy, the effect is mainly on
investment, while in the closed economy, because savings equals investment in equilibrium,
Turkan Aliyeva Karimova
the effect is mainly on impact on the real interest rate and on investment and savings in the
long-run.
The effect on investment in the small open economy, when not accompanied by an
increase in savings, is immediate, while the effect on investment and savings in the closed
economy is a slow one over time, leading to a transitional period when instead of capital in
ows, the economy suffers a higher real interest rate over time. We show that inequality
increases for our calibration, but also that the intuition leads to this being a more general
result.
Their conclusion is that the elimination of the capital income tax, when compensated
by an increase of the tax on labor income, decreases the welfare of the households in the left
side of the welfare distribution, that is the poorest of the economy. As the poorest households
have labor income as the main source for fi nancing of consumption, if wages would increase,
as in Harberger (1995) with the elimination of capital taxation, it may not be the case that they
would suffer in terms of welfare. This is the literature dilemma that this article tries to clarify.

Empirical Strategy

Being the main objective of this paper to understand the distributional effects of the
elimination of capital taxation we begin by describing in this section the roots of
heterogeneity at the time the reform is implemented.
After we will discuss how we can compare the cross section distribution of welfare,
which depends on these individual characteristics as well as on equilibrium prices, before and
after the fiscal reform. Being the main objective of this paper to understand the distributional
effects of the elimination of capital taxation we begin by describing in this section the roots of
heterogeneity at the time the reform is implemented. Households are heterogeneous in labor
efficiency and in non-human wealth. Each household i has a deterministic labor eficiency
level measured by Ei ;which is constant overtime. This same household holds at every time
period, t , a stock of non-human wealth, Ait; which is decompose in every period in physical
capital, Kit; domestic bonds, Bit and, if the economy is not closed, external assets B it; being
this decomposition chosen in the previous period, t 1. At time 0 this individual non-human
wealth, Ai0; is exogenous and its distribution, jointly with the distribution of labor effiency
levels, Ei ;characterize the sources of heterogeneity in this problem. Therefore we assume that
agents are identical in every other characteristic.
There is just one good produced in every period, this good is identical to the one
produced in the rest of the world and there are no restrictions to the tradability of this
good.The model represents a small open economy with perfect capital mobility, that is an
Turkan Aliyeva Karimova
economy integrated in a global capital market. It is a real economy in the sense that we
abstract from money as a facilitator of transactions.
As described in Correia (1999), comparison of distributions can be very simplied
when agents are heterogeneous but Gorman ggregation is still possible. Most used
preferences fall under the class that allows for aggregation. Given cross section empirical
evidence, we propose the type of preferences used in Greenwood, Hercowitz and Huffman
(1988) (GHH), which are characterized by a zero wealth effect on labor decisions.
This characteristic implies that households with higher stocks of financial wealth work
the same ammount of hours than poor households, when having the same labor eficienc.
We also assume that fundamentals in the rest of the world are identical to the ones of the
small economy. These assumptions imply that, with no costs of adjustment of capital, the
economy will converge immediately to the new stationary state, following the change of
policy.

As the this article is to understand the different effects that the change of policy can have on
different households that live in the small open economy, it is important.
The objective of this article is to determine in which conditions the elimination of the
tax on capital income, when compensated by an increase of the labor tax, improves equity in
the small open economy. Where the economy is characterized by a constant positive tax rate
on capital, with the alternative situation, policy , where the economy is characterized by a
zero tax rate on capital.
Therefore, in this environment, the effect of the elimination of capital taxation on
equity depends completely on the roots of households heterogeneity. Then the question
proposed is an empirical one: what is the root of the households heterogeneity observed in
most industrialized or emerging countries? Cross section data tells us that both wealth and
earnings are not equally distributed across households. If we , use general characteristics of
empirical evidence, that the joint distribution of those two household' characteristics, labor
efficiency and initial wealth, satisfy .
We can say that in the open economy the elimination of capital taxation leads to an
immediate increase of the capital stock. We can say that it has an immediate investment effect
and a slow saving effect since the investment is nanced by external savings. This increase will
ceteris paribus lead to a negative effect on interest rates and to a positive effect on gross
wages. In the closed economy savings and investments should be equalized at every period.
The increase of demand for investment without having enough savings, since it is costly to
decline consumption, leads to an increase of interest rates.
Turkan Aliyeva Karimova
Therefore the investment and savings increase slowly to achieve the new capital labor ratio at
the new steady state. We can say the equilibrium is achieved immediately in the small
economy through increase in quantities and in the closed economy by increase in prices.
We confirm that for the cross s of the US and for the calibrated model the welfare of
those households decline with the elimination of capital taxation . The obtained decline of
welfare for the poor of the economy confirm the result in Domeij and Heathcote (2004) and in
Garcia-Mila et al (2010) that the the elimination of capital income declines total welfare
because it declines the welfare of the poorest households in the economy. As said the method
used by those authors differ from ours because they use non-agregable preferences and/or no
heterogeneity in labor eficiency, when there are no idiosyncratic shocks. At Garcia-Mila et al
(2010) the equilibrium prices are dependent on the proposed joint distribution of labor
eficiency and initial wealth. Here we show that even if this is not the case in our model the
qualitative results are identical. This can be read as being ours a simpler method, and
simultaneously a good approximation for the results, or that the distributional effects on the
equilibrium aggregates are for this model of second order of importance.
The environment in which the exercise of corparate tax income was developed was
the one of a small open economy with perfect capital mobility, while Garcia-Mila et al.
calibrate their model for the US, which is described as usual by a closed economy model.
How does this change of environment revert the results on equity in a such a strong way? The
fundamental difference is that in the environment described until now the real interest rate
was exogenous to policy. That is it did not react to the elimination of capital taxation. While
in Garcia-Mila et al. the real interest rate is a variable that reacts to policy, due to the changes
in saving and investment associated to the change in taxation. The same would occur if, even
when considering a small open economy, we would assume that the rest of the world,
composed by a set of identical small open economies, was changing policy in a similar way
and simultaneously to the specifi c small open economy under study.
How should it change when the economy is represented by a closed economy, that is
one economy where the path for the real interest rate would react to the change of policy. The
environment is identical to the one developed , except for capital immobility and goods
nontradability which implies that, in every period, market clearing imposes that the sum of
private consumption, public consumption and investment has to be equal to the production
realized in the economy. This change, which is equivalent to the real interest rate being
endogenous to policy in this economy, implies that, contrary to the former model which was
analyzed analytically, now we have to use a numerical solution method for the computation of
the equilibrium.
Turkan Aliyeva Karimova

Conclusion .

We show in this article that the effect on equity of the elimination of the tax rate on
capital income depends in a crucial way on the globalization of capital markets. Meaning
that whether the elimination of capital taxation leads to a change in the path of the real
interest rate or to ncapital in ows into the country makes the whole difference for the result.
when compensated by an increase of labor taxation, depends in a crucial way on the effect
that the change of policy has on the path of the real interest rate.
When we analyzed a small open economy where that rate is exogenous to policy the
result is that the poorest households of the economy increase welfare as a result of the change
of policy.In a closed economy the elimination of capital taxation leads to an increase of
inequality trough the change of the path of capital labor ratio and the eect of this on the net
interest rates; in a small open economy the unilateral decision of its government to eliminate
the tax on capital implies a capital in ow that lead to a real interest rate always equal (except
in period zero) to the limiting one of the closed economy. It is the capital in ow in opposition
to the change of the net return that implies the increase of the sum of the present value of net
wages in the first case and the opposite in the second one.
The result obtained for the closed economy can occur either because the domestic
capital market is segmented from the international market, or because, being capital markets
open internationally, the change of policy is taken simultaneously by every other country.
Also in this case the adjustment is done through changes in the net international rate of
interest and, in the limit when countries are identical, there is no immediate movements of
capital across countries but just a change of the interest rate. capital in each country will
increase slowly over time as in the closed economy. Theoretically the effect on equity would
also depend on the roots of heterogeneity across households. However the advantage of our
method is to be able to guarantee that the result is well defined: worsens inequality when the
country is a closed economy and improves when the policy change is realized in a small open
economy.
The important characteristic on data is the robust characteristics across economies that
wealth is more unevenly distributed across households than earnings. Besides, as well
established in the literature, the effect on effiency of the elimination of capital taxation is
positive, both for the closed and for the small open economy. Both effects, this one on
effiency and the one on inequality, imply that the decision to implement that policy leads to
an increase in the welfare of the poorest households in an economy where the change of
Turkan Aliyeva Karimova
policy does not alter the real interest rate, that is in the small open economy. On the contrary
the segmentation of capital markets in the closed 25 economy can turns this result round.
In contrast, study characteristics related to globalization measures give rise to totally
different findings concerning the issue under examination. Moreover, we provide empirical
evidence that several other study characteristics (e.g. whether a particular study has been
published in a political or in an economic journal) do have systematic impact on the reported
results. These conclusions are very important for future empirical studies examining the
effects of globalization on capital taxation.