THE KHALDUN-LAFFER CURVE REVISITED: A PERSONAL INCOME TAX-BASED ANALYSIS FOR TURKEY.

• Author: Sen, Huseyin

1. Introduction

One of the most important and well-known arguments of supply-siders is the notion of the Khaldun-Laffer curve (henceforth the K-L curve) (1); based on persistent high tax rates in the US, supply-siders have claimed that lower tax rates would mean higher tax revenue and that consequently higher rates would result in lower tax revenue. Hence, reducing current tax rates would be a good tax policy because it will not only maximize tax revenue but also motivate economic activity. This argument of supply-siders made the K-L curve popular in the early 1980s. The K-L curve is the graphical representation of the aforementioned postulate. It depicts a classic bellshaped relationship (2) between tax rates and tax revenue: as the tax rate increases the tax revenue increases till it reaches an optimal point of maximum tax revenue; any further increase in the tax rate decreases the revenue due to the disincentive effects created by higher taxes. Laffer (1981, 2004) argues that higher tax rates will discourage work and production by removing incentives, triggering sluggish growth, and thereby lowering tax revenue collected by the government.

As with many other developing countries, maximizing revenue is important for Turkey in order to realize its fiscal targets. Considering the undesired consequences of the alternative ways to finance budget deficits (especially, borrowing and printing money), the significance of having an ideal tax system in which tax rates generate maximum tax revenue for the government without inducing a dampening effect on the economy has been well understood.

Turning to the case of Turkey, this paper proposes to revisit the K-L curve. In this context, the paper first endeavors to examine the validity the K-L curve empirically and then attempts to find the optimal tax rate at which the maximum amount of tax revenue can be collected. The paper also aims to: (i) investigate whether present tax rates are lower or higher than the revenue-maximizing tax rates; (ii) determine the income and substitution effects of tax rates and thereby design ideal tax rates for fiscal policy purposes; (iii) avoid the undesired consequences of taxation incurred from high tax rates (i.e. tax avoidance, tax evasion, preferring leisure to work, changing the scope and/or structure of economic activity, etc.) which result not only in a reduction of tax revenue but also negatively affect economic activity and thus growth, and (iv) contribute to the limited empirical literature related to the K-L curve for developing countries like Turkey.

Since the curve theoretically establishes a non-linear link between tax rates and tax revenue, we have chosen the personal income tax (hereafter, the PIT) as the case for testing the validity of it. The reasons for doing so, first and foremost, is that the PIT is a tax that is levied on individuals' taxable income. Second, this tax accounts for the largest share of the central government's tax revenue after VAT and the special consumption tax. Thirdly, the PIT is a tax instrument that is well-suited to examining the K-L curve.

The PIT is one of four important taxes that generate the highest tax revenue in the Turkish tax system. The other taxes are the corporate income tax, the value-added tax, and the special consumption tax. These four taxes account for an average of more than 80% of total central government tax revenue. Although two indirect taxes (the value-added tax and special consumption tax) dominate the tax system and constitute nearly 65-70% of the central government's tax revenue on average, the PIT also has a significant place in the tax system, accounting for almost 20% of it. Moreover, it is the sole progressive tax (if wealth tax, which constitutes a minor portion of total tax revenue, is ignored). At present, the PIT contains four brackets with rates of 15, 20, 27 and 35%, respectively. As a rule, these rates are not frequently changed. The latest alteration was made in 2006 when the number of tax brackets was reduced from 6 to 4. However, at the end of the year, before starting the new fiscal year, tax brackets are re-adjusted for subsequent fiscal years in line with developments in inflation of the current year.

The remainder of the paper is organized as follows. Section 2 outlines the theoretical background and related literature. Section 3 describes the data and econometric methodology, while Section 4 presents the estimation strategy and results of the paper alongside their interpretation; the final section concludes.

2. Theoretical background and related literature

2.1. Theoretical background

While Laffer was discussing the US President's proposal for tax increases in a dinner meeting in 1974, he charted a classic bell-shaped curve on a cocktail napkin, illustrating the trade-off between tax rates and tax revenue (see Laffer, 1981, 2004; Wanniski, 1978, 2005). Soon after, Wanniski (1978) named this trade-off as 'the Laffer curve'; however, Laffer (1981, 2004) notes that the idea behind the curve belongs to Ibn Khaldun, a 14th-century Muslim philosopher. Giving a special reference to the opus of Khaldun, the Muqaddimah, Laffer draws attention to the background of the curve with the words of Khaldun himself: 'it should be known that at the beginning of the dynasty, taxation yields a large revenue from small assessments. At the end of the dynasty, taxation yields a small revenue from large assessments' (Khaldun, 1980 [1377], p. 80).

However, contrary to the arguments above, some authors, such as Fullerton (1982), van Ravestein and Vijlbrief (1988), and Hsing (1996), attribute the roots of the curve to Adam Smith, an 18th-century Scottish economist, referencing his words: 'high taxes, sometimes by diminishing the consumption of the taxed commodities, and sometimes by encouraging smuggling, frequently afford a smaller revenue to government than what might be drawn from more moderate taxes' (Smith, 2008 [1776], p. 835). Despite these counter-arguments, we prefer to label the curve as the K-L curve to give credit to both Khaldun and Laffer.

The curve exhibits a one-way causality from tax rates to tax revenue, illustrating a hump-shaped relationship between tax rates and tax revenue. The upward-sloping side of the K-L curve displays the normal range, whereas the downward-sloping side shows the prohibitive range (Wanniski, 1978; Laffer, 1981, 2004). This implies that a given tax revenue can be collected with two different tax rates, one falls into the normal range of the curve, and the other one falls into the prohibitive range; the rational fiscal policy option would be choosing the lower rate, which falls into the normal range because the income/incentive effect dominates the substitution/disincentive effect of taxation.

The K-L curve is still being discussed, although more than three decades have passed since its postulation by Arthur Laffer. The focal point of discussion is whether the K-L curve is true or false, or only an approximation. The discussions are centered specifically on whether there is such a relation between tax rates and tax revenue, as emphatically argued by Laffer (1981, 2004). If so, what is the revenue-maximizing tax rate, i.e. the optimal tax rate?

The K-L curve received a great attention especially in the early 1980s with the arguments of the prominent supply-siders such as Paul Evans, Michael Boskin and Martin Feldstein, asserting that 'lower tax rates would mean higher revenue because existing rates were too high to maximize tax revenue, - that is, tax rates were so high that fewer taxed goods were being produced and the overall effect was lower tax revenue' (Becsi, 2000, p. 53). Even today, the curve still receives remarkable attention in tax-related discussions, especially in the United Stated of America, where a conservative political movement within the Republican Party that is called 'Tea Party' keeps the topic always on the agenda and makes the reduction of taxes one of its primary objectives (Tanzi, 2014).

Despite these facts, the curve encounters severe criticism in the literature, such as: (i) the K-L curve, in the words of Tanzi (2014), is a wonderful propaganda device for conservatives; (ii) the idea that the Laffer curve has disincentive effects on high tax rates is not precise, and (iii) there are suspicions about the existence of a relationship between tax rate and tax revenue. More specifically, the curve that establishes a classic bell-shaped relationship between tax rates and tax revenue obtained by the government may not reflect the reality. Depending on a number of factors, such as taxpayers' reaction to the taxes, the openness degree of the economy, efforts of the tax administration, structure of labor market, and etc., the curve may take different shapes, rather than proposed by Arthur Laffer (3).

In the literature, we see at first glance that there have been many studies on the K-L curve, a large majority of which have focused on the theoretical aspects of the curve, rather than empirically exploring it. There are also numerous counter-studies such as Henderson (1981), Sanyal, Gang and Goswani (2000), Busato and Chiarini (2013), which critically assess the curve in terms of different aspects, ranging from its assumptions to its validity.

This paper contributes to the literature not only by focusing on the PIT of Turkey in the context of the K-L curve in the long run, but also by including control variables...