Does Informality Hinder Self-Enforcement? Evidence from Value-Added Tax in Pakistan

One of the most significant tax policy developments of the last thirty years has been the adoption of VAT by a large number of developing countries. These countries had traditionally been reliant on trade taxes (import duties) for their revenue needs. With the decline of trade taxes in the last two decades of the twentieth century, due mainly to the multilateral trade liberalization agreements, these countries had to look for alternative domestic sources to replace the lost revenues. Since a broad-based VAT was generally considered to be the most suitable alternative, the tax was adopted rapidly in more than 130 countries including almost all developing and transition economies. For example, the number of African countries with a VAT increased from 2 to 30 in the 1990s.

Recently, however, doubts have started to emerge if the replacement of trade taxes with VAT in developing countries was a particularly good tax policy considering that a large part of economic activity in these countries takes place in small firms who are extremely difficult to monitor and register. Two recent studies have shown that on average low-income countries were able to recover only a fraction of -- less than 30% -- lost trade revenues through domestic sources. More importantly, such recovery was not significantly stronger in countries who adopted VAT compared to those who did not. In addition, the theoretical results which prove VAT to be superior relative to other alternatives hold only if the government has no difficulty in registering all firms in the economy.

Whether and to what extent VAT is suited to the environment in developing economies is ultimately an empirical question. There, however, is still very little evidence on the performance of the tax in such settings. The purpose of this study is to provide such evidence from the experience of one representative economy, Pakistan. There are a number of reasons that make Pakistan an ideal country to research this question. Pakistan has a long history of implementing VAT, introducing it more than two decades ago in 1990. It is also one of the countries where efforts to replace trade revenues with VAT resulted in a substantial decrease -- from 15% to 9% of national output -- in overall tax collection. Furthermore, Pakistan has a large informal sector estimated currently to be around 40% of the official economy. The country, thus, shares important features with a large group of other developing and emerging economies who have had a similar difficult experience with VAT. The results of the research, therefore, will be generalizable and of interest to policy makers in a broad range of countries.

One of the more exciting features of the Pakistani context is the nature of data available for the research. Pakistan has available in digitized form returns filed by all registered taxpayers since the introduction of the tax. More importantly, since 2006 it has been mandatory for all registered taxpayers to file a statement containing transaction-wise details of sales and purchases made by them. The availability of this data is unusual and extremely helpful as it will enable us to create a linkage between firms in various parts of a supply chain, which is vital for studying efficiency and compliance questions associated with VAT in developing country settings. In addition, the data also includes income tax returns filed by all corporate and non-corporate firms in Pakistan since 2006. The availability of income tax returns is useful because the exemption cutoff for income tax and VAT are set at different levels. The lower income tax cutoff will help us see how firms who are too small to register for VAT get affected by it. The novel dataset spanning a number of years in combination with very rich variation created by a series of tax changes implies that the proposed research will be in a unique position to generate very useful insights for tax policy in developing countries.

The proposed research will benefit a large number of developing countries who are struggling to raise significant revenue from VAT and as a consequence are unable to develop their fiscal capacities. To ensure that these countries are able to extract maximum benefit from the research, we will engage with the policy institutions, especially the IMF and World Bank, to ensure that the results reach the policy community in these countries. This will be achieved by presenting our work at seminars in these organisations and by engaging directly with the researchers who formulate policy advice for developing and emerging economies in the areas of development and public finance. Given that a large number of target countries still rely on advice from these organizations, such engagement is expected to lead towards the adoption of the policy recommendations in a broad range of countries.

In addition to working with the multilateral institutions, we will seek direct interaction with policy makers in developing and emerging economies. For this purpose the University of Manchester's existing network with developing countries through its research centres such as the Brooks World Poverty Institute and Effective States and Inclusive Development will be used. We will also present our work at yearly conferences organized by the International Growth Centre, which attract a large participation of tax policy advisers and practitioners from around the world. The benefits of our research will reach the targeted recipients through these networks as well.

The results of the research will be of specific interest to the Government of Pakistan and in particular its tax administration, the Federal Board of Revenue (FBR). We will work closely with the FBR to see that major policy recommendations of the project get implemented. What will facilitate this process is the fact that I have worked for the FBR for more than ten years and have an extensive network of ex-colleagues and friends in the organisation. I also have a very successful track record of getting policy recommendation of my earlier research incorporated into the tax code. In order to build on the past success and to maximise project impact the following plan will be undertaken. Before the start of the project, a workshop will be organized at the FBR to seek the organisation's perspective on the research questions. During the meeting, a structural arrangement for effective coordination of the research with the FBR will also be worked out. This two-way exchange of ideas will be maintained throughout the duration of the project. Following the analysis of the research data, the FBR and other research and policy making institutions in Pakistan will be engaged again. Through this engagement a consensus plan for implementing recommendations of the research will be evolved. Given that stagnant revenue from VAT is one of the most salient tax policy concerns in the country the project is expected to receive very keen attention from the policy community in Pakistan.

Another group to benefit from the proposed research will be the institutions who give tax policy advice to developing countries such as the International Monetary Fund. There are growing concerns that some of the policy advice offered by such institutions is based on theoretical results which do not hold in developing country settings (see for example Best et al. Forthcoming). The proposed research will benefit these institutions by enriching their knowledge on tax environment in developing countries and by helping them tailor their policy advice according to the realities in these countries. The proposed research will also be of interest to development focused policy institutions, such as the World Bank, who help developing countries devise policies to facilitate the transition of informal firms to formal firms.