A political economy of tax havens
3.2 The Model
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
26
such an international cooperation is beneficial to the non-haven countries (e.g., Haufler and Runkel 2009; Slemrod and Wilson 2009). However, we demonstrate that, in the presence of lobbying, the cooperation on reducing the international tax planning activity can reduce the non-haven countries’ welfare. The reason for this result is based on two findings. First, the equilibrium capital tax rate is lower than the level that maximizes a non-haven country's social welfare. Secondly, a coordinated reduction in the international tax planning activity can lead to a lower tax rate.
Combining these two outcomes shows that the cooperation on reducing international tax planning activity can cause the sub-optimally low tax rate to deviate further below the efficient level, and thus reduce all non-haven countries’ social welfare.
The remainder of this chapter proceeds as follows. Section 3.2 describes the basic tax competition model with tax havens. Section 3.3 introduces the capitalists’
lobbying and obtains the political equilibrium capital tax rate. Section 3.4 and 3.5 examine, in the presence of political power, how international planning affects the equilibrium capital tax rate and social welfare. Section 3.6 discusses the role of workers' lobbying. The last section concludes.
3.2 The Model
Our basic model builds on the works of Zodrow and Mieszkowski (1986) and Haulfer and Runkel (2009). We consider a large number of identical (non-haven) countries. Each country contains three types of residents: capitalists, workers, and pensioners. All residents are immobile across countries and the residents of the same type are identical. We normalize the number of the pensioners to unity, and let nk and nl be the numbers of capitalists and workers, respectively.
Each capitalist possesses h units of immobile capital, kn, and one unit of internationally mobile capital, km. Each worker is endowed with one unit of labor,
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
27
and inelastically supplies labor to firms. Finally, pensioners, defined as those who are endowed with nothing, receive transfer payments from the government.
Suppose that each country contains a firm, which is a price taker in the international market. The firm hires the immobile capital, mobile capital, and labor to produce an output. The two types of capital are perfectly substitutes in producing the output.
In addition to the non-haven countries, there also exist several countries or jurisdictions levying no taxation on capital,14 which we refer to as tax havens. The existence of the tax havens enables the firms to exert international tax planning. More specifically, a firm can set up a financial subsidiary in a tax haven. The subsidiary does not produce output;15 it only makes an intra-company loan (usually merely paperwork) to its parent company, whose country permits the deduction of interest payment for this loan. As shown below, the firm is able to reduce overall tax burdens by using the internal debt financing.
In reality, these intra-loan strategies are often constrained by a number of factors, including transaction costs, the agency problems, long-term natures of tax codes, or the numbers of available tax havens. Most of these factors are beyond the control of the firms.16 To characterize the tax planning in the simplest way, we follow the specification of Hong and Smart (2010) by assuming that the a firm’s internal debt to its subsidiary is bounded by an exogenous proportion, [0,1). Also by following Hong and Smart (2010), we assume that the firms will issue the debt up to the upper bound.
Accordingly, the profit function of the firm is written as:
14 Allowing tax havens to impose a positive tax rate on capital income does not alter the following results.
15 According to the empirical evidence, tax havens are usually very small jurisdictions (Hines and Rice, 1994; Dharmapala and Hines, 2009). Without loss of generality, we neglect productive activities in tax havens.
16 See Desai et al. (2004) and Hong and Smart (2010) for more discussions.
‧
respectively. The government imposes a source-based tax at rate of t on each unit of capital. The capital tax rate is restricted to be non-negative. We note that the international tax planning lowers the effective tax rate on the mobile capital to(1)t. The wage rate is denoted by w.
The firm chooses kn and km to maximize (3.1). The first-order conditions for profit maximization are:
t
Since the mobile capital encounters a lower effective tax rate, its net return is higher than the immobile capital’s.
In addition, the wage rate is adjusted to the point where it is optimal for the firm to employ all labor supply. Inserting (3.2) and (3.3) into (3.1) gives the aggregate regards the net rate of return of the mobile capital as given. Therefore, it follows from (3.3):
17 For notational simplicity, the fixed labor input is omitted in the production function.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
29
Equation (3.5) says that the firm’s capital demand decreases with t. We note that the adverse effect of t on the capital demand decreases with . It is because with a larger
, the increase in the user cost of mobile capital due to an increase in t is lower.
Hence, the reduction in the firm's capital demand is smaller with a higher level of tax planning. Equation (3.6) indicates that a rise in international tax planning lowers the user costs of mobile capital, and therefore increases the capital demanded.
Although an individual country’s policy cannot change the net return on the mobile capital, it affects the net return on the immobile capital, rn. The effect of t on
rn is given by drn
dt (3.7) Equation (3.7) shows that the effect of t on rn depends on the level of tax planning;
the larger , the larger adverse effect of t on rn will be. From (3.2), the amount of capital employed and the tax rate jointly determine rn. The capital outflow due to an increase in t raises the marginal product of capital, which in turn mitigates the adverse impact of t on rn. However, with a larger , an increase in t causes less capital outflow, leading to a greater negative impact on rn.
A representative capitalist’s preferences are given by:
uk h rn rm.
A representative worker’s preferences are described by:
( ) ( )
l l
u w f k k f k n .
Finally, the utility function of a representative pensioner is given by (recall that the number of the pensioners is normalized to unity):
[(1 ) ]
p k
u t kn h .
The social welfare function W is defined as a weighted utilitarian function:
(1 )
k l p
W W W W , (3.8)
‧
pensioners’ welfare. For the redistribution concern, the capital tax revenues are distributed to the pensioners in a form of lump-sum transfers.Each type of residents’ aggregate welfare is given as follows:
( ) of each type of residents:
attracts capital inflow, which in turn enlarges the wage rate. An increase in has a similar effect on the workers’ welfare as a reduction in the tax rate. We also notice the adverse effect of on the capitalists’ welfare. Although a higher increases the net return to mobile capital, it also attracts capital inflows, which reduce the returns to both immobile and mobile capital. Equation (3.9b) reveals that the latter effect outweighs the former one, so that an increase in makes the capitalists worse off.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
31
Finally, both the effects of tax rate and tax planning on the welfare of the pensioners are ambiguous.