ABSTRACT

Despite several major tax policy reforms since independence and implementation of a series of important new taxes in recent years, as documented in Chapter 2, tax policy in Mozambique has received only limited analytical treatment. Furthermore, where studies have taken place, these have tended strongly to focus on tax revenues from a government and therefore macroeconomic standpoint. As such, while government policy ostensibly seeks to simultaneously provide a sustainable source of revenues and promote economic growth, there is very little formal analysis of how tax policy is applied in practice or of its subsequent impact at the microeconomic level. This chapter addresses this lacuna by analysing enterprise data relating to tax burdens and undeclared output using results from two manufacturing enterprise surveys carried out in Mozambique in 2002 and 2006. Government tax policy must find a balance between revenue, equity and

efficiency by raising sufficient revenue, taxing individuals and firms in an equitable manner, minimizing the effects on incentives and administrative costs, and promoting stability and economic growth. Individuals and firms then weigh up the benefits of tax evasion against the costs of detection and punishment, behaving according to the outcome of that trade-off. As Burgess and Stern (1993) point out, developing country economies are generally characterized by a number of factors which hinder governments in the implementation of an effective tax policy, including a large primary sector, economic and social dualism, extreme income inequality, a concentration of economic activity in very small enterprises, extensive uses of permits, licences and rations, weak administrative capabilities and pervasive corruption. As such, the cost of non-compliance is often relatively low, resulting in higher levels of evasion. In addition, a lack of productive investment often leads developing coun-

try governments to use tax incentives to encourage investments, thus further eroding the tax base. As a consequence, evidence suggests that in developing economies even broad-based taxes only manage to capture revenues from a small proportion of the statutory tax-base (Gordon and Li 2005), resulting

in what can be termed revenue concentration. Gauthier and Gersowitz (1997) and Gauthier and Reinikka (2001) analyse tax-base erosion and concentration for Cameroon and Uganda, respectively, and find that this is strongly associated with firm size, defined by the number of employees. They find an inverted-U relationship between tax burden and firm size, with firms at the lower and upper ends of the firm-size distribution experiencing a relatively low tax burden compared to those in the middle of the size distribution. This is interpreted as a result of small firm tax evasion, large firm access to tax exemptions and a lack of means to escape for firms in the middle of the size distribution. This revenue concentration violates principles of equity and efficiency,

with potentially negative economic consequences. By creating an institutional environment which varies according to firm size and/or other firm characteristics, incentives for enterprise growth are potentially distorted, with negative consequences for economic growth, formal employment creation and ultimately poverty reduction. In addition, incentives to comply with tax laws are reduced by a perception that few others are doing so, such that tax revenues themselves are affected, potentially resulting in a vicious cycle of higher and more distortionary taxes and further rounds of revenue concentration on more visible firms to compensate for losses of government revenue from economic activity which goes unreported.1