REF: 2014

Impact case study

Helping countries improve their tax systems


The Government implemented a major reform of income tax that followed the key recommendations of an LSE paper, which was described in an International Growth Centre Policy Brief that was circulated immediately prior to the Parliamentary debate on the reform.

Henrik Kleven

Research by

Henrik Kleven


Download (128.5 KB/PDF)

LSE and IGC researchers helped countries as diverse as Denmark and Pakistan to improve the design of their tax systems

What was the issue?

A failure to collect all the tax revenue owed to the Government means there is less money to spend on infrastructure and health and education services that are vital to economic growth.

A well-designed tax system can motivate positive changes in the way people behave, such as attracting more people into the labour market, increasing the amount of hours worked, and encouraging reporting and payment of taxes owed. Specifically, an efficient tax system will be designed to provide incentives that encourage people to get a job, rather than to claim benefit, and to increase the amount of hours they work. It will also be designed to remove any loopholes that taxpayers can use to avoid paying what they owe.

What did we do?

Researchers from the LSE, led by Professor of Economics Henrik Kleven, decided to look at the tax systems of two countries, Pakistan and Denmark, and identified specific and significant failings in each system.


It is typical in tax systems that the rate of income tax people pay increases as incomes cross certain thresholds. However, whereas countries such as the UK apply the higher tax rate only on the income above that threshold, in Pakistan the higher tax rate applied to an individual’s entire income, producing a sudden jump in their tax liability.

The LSE researchers found that this led to a bunching of incomes below each threshold, resulting in lost earnings of 10-15% as people preferred to stay on the below-threshold incomes. The researchers also showed that changing how these ‘backward-looking’ thresholds worked – so that only the income above the threshold was subject to the higher tax rate – would have few negative consequences on the efficiency of the system or on how people behaved.


In Denmark the researchers found that rates of tax evasion were much higher among earners who reported their own tax liability compared with those whose earnings were recorded by third parties such as their employer, bank or investment fund manager. In particular, they found capital gains tax liabilities were particularly vulnerable to evasion because of the absence of third-party reporting by financial institutions on the prices at which shares were sold.

The researchers also looked at the best way of taxing married couples. They found that when the income of the secondary earner – typically the wife – goes up, the marginal tax rate paid by the primary earner should go down. (The marginal tax rate is the amount of tax paid on an extra pound, rupee or kroner earned.) They showed that this could be achieved by combining means-tested transfers based on the overall family income, with income taxes based on individual incomes – as in the UK system. This strikes the best balance between the goals of encouraging labour supply and economic efficiency and ensuring a fair distribution of after-tax incomes across families.

What happened?


The Government implemented a major reform of income tax that followed the key recommendations of an LSE paper, which was described in an International Growth Centre Policy Brief that was circulated immediately prior to the Parliamentary debate on the reform. (The International Growth Centre (IGC) is a joint venture of LSE and Oxford University.) The LSE work was the only existing study of the distortionary effects of Pakistan’s tax system.

The paper was discussed at a meeting of the Economic Advisory Council to the Prime Minister of Pakistan, and one of the LSE researchers travelled to Pakistan to discuss the findings with the country’s Tax Policy Secretary.

The reform changed the ways the thresholds worked so that higher tax rates were only paid on incomes above each threshold. It also reduced the degree to which self-employed workers and wage earners were taxed differently.

Following the success of this reform, Pakistan’s Federal Board of Revenue established a more formal framework with a team of researchers at LSE and IGC, led by Professor Kleven. This has enabled researchers to undertake research projects of direct interest to Pakistani policymakers, who have given them unprecedented access to FBR administrative data and institutional know-how. The two sides have held conference calls to discuss tax design and reforms that can increase revenue, efficiency and growth in Pakistan.


The Danish Government implemented a major tax reform in 2010 that expanded third-party information reporting by requiring financial institutions to report the sales prices of financial assets to tax authorities. The design of the tax reform was based on the recommendations of a government-appointed Tax Commission (Skattekommissionen), which included one of the researchers (Claus Kreiner) involved in the LSE research project on tax evasion in Denmark.

The reform package also changed the tax treatment of married couples by abolishing a particular income tax that was based on family rather than individual income. This resulted in tax liability being wholly based on the individual with welfare transfers based on the total earnings of the family. The Skattekommissionen cited the LSE research in its recommendations to the Government.

Search all impact case studies.

More by this author