Tax policies that may work best for Pakistan
When politicians don’t pay taxes, it evokes much anger
The writer is a postdoctoral fellow at Stanford University and a non-resident fellow at the Lahore-based Consortium for Development Policy Research
When politicians don’t pay taxes, it evokes much anger. But the problem is not just politicians: it’s all those who should pay taxes but don’t. Pakistan’s tax-to-GDP ratio is only 11 per cent, one of the lowest in the world. Countries with similar levels of development have ratios around 15 per cent. While many people estimate that Pakistan’s tax base should consist of three to four million people, only a million Pakistanis file their tax returns or appear on employer tax statements.
The main problem is that Pakistan’s current tax policy assumes that the government has both good information on the tax base and the capacity to collect taxes at little cost. But Pakistan faces major challenges on both counts, leading to widespread evasion and inadequate collection of taxes. This makes it harder for the government to pay for vital infrastructure and social development projects to raise incomes, reduce poverty and address inequality.
Reforming the tax administration is one way forward, but it is difficult and will take time. A growing body of evidence suggests there is another way: Focus on innovative methods that suit Pakistan’s ground realities and that the current administration can use to collect more revenue now.
To understand this approach, consider three tax collection scenarios.
In the “first best” scenario, governments have perfect information on taxpayers and can collect taxes effectively at no cost. What is important here is that they know how much potential income citizens have. With that information, governments can spread the burden of collecting taxes in an equitable and efficient way. While this is ideal, no country in the world has a “first best” setting because it is impossible to know what taxpayers’ potential income is.
In the “second best” scenario, governments have the capacity to collect taxes, but information is weaker. Governments only know how much people earn — their actual income — but not their potential income. So they tax according to how much people really make: Higher earners are taxed more and lower earners are taxed less. This yields less revenue than a tax based on potential earnings because higher earners can lower their incomes to pay less taxes. But it is the best countries with relatively effective tax administrations can do.
In the “third best” scenario, the government neither has the capacity to know even actual incomes nor to collect taxes. Taxpayers may not truthfully report their incomes and tax administrations cannot collect taxes without great effort and cost. This scenario is closer to reality, especially of lower income countries like Pakistan.
Pakistan’s current tax policy operates as if it were in a “second best” world, but it better fits into the “third best” category. This has real consequences for how much tax revenue can be collected.
Take the example of corporations. Right now, Pakistan taxes corporate profits. This would be an effective way to tax in a “first” or “second best” setting because the government would know how much actual profit the corporations make and would tax them. But with the current weak system of publicly available company accounts and the inability to verify the accuracy of corporate profits, corporations have a strong incentive to under-report profits and pay fewer taxes.
If Pakistan took a “third best” approach, what would it do instead? One solution is a minimum tax on corporate revenue. Pakistan already does this for corporations whose profit is too small to tax.
With the help of collaborators, I conducted a study of the minimum tax and found that it prevented 60 to 70 per cent of the profits of the corporations in the study from going misreported. This evidence implies that a “third best” tax policy — in which the minimum tax is implemented more widely — would leverage the revenue-raising power of the minimum tax while minimising its distortionary costs to the economy. Implementing this would not entail a strenuous restructuring of the entire tax administration. This is a policy change that could be done easily with existing tax capability.
Once the government recognises that it cannot observe people’s income and trust them to report directly, it opens up more areas to apply a “third best” approach. For example, using consumption behaviour and salaries as alternative sources of information to widen the tax net would make greater sense.
The withholding tax — which is levied in Pakistan through a tax deducted at source from a payment of income — demonstrates that, to a certain extent, this is already acknowledged. It has been somewhat successful in Pakistan because the government only has to monitor employers who act as the government’s withholding agents, and employee salaries are less likely to be misreported to them.
Reforms to improve the capacity of tax administration to collect information and levy taxes will undoubtedly be important in the long term. But to support Pakistan’s development agenda now, the smartest approach could in fact be the “third best”.
Published in The Express Tribune, January 7th, 2017.
The main problem is that Pakistan’s current tax policy assumes that the government has both good information on the tax base and the capacity to collect taxes at little cost. But Pakistan faces major challenges on both counts, leading to widespread evasion and inadequate collection of taxes. This makes it harder for the government to pay for vital infrastructure and social development projects to raise incomes, reduce poverty and address inequality.
Reforming the tax administration is one way forward, but it is difficult and will take time. A growing body of evidence suggests there is another way: Focus on innovative methods that suit Pakistan’s ground realities and that the current administration can use to collect more revenue now.
To understand this approach, consider three tax collection scenarios.
In the “first best” scenario, governments have perfect information on taxpayers and can collect taxes effectively at no cost. What is important here is that they know how much potential income citizens have. With that information, governments can spread the burden of collecting taxes in an equitable and efficient way. While this is ideal, no country in the world has a “first best” setting because it is impossible to know what taxpayers’ potential income is.
In the “second best” scenario, governments have the capacity to collect taxes, but information is weaker. Governments only know how much people earn — their actual income — but not their potential income. So they tax according to how much people really make: Higher earners are taxed more and lower earners are taxed less. This yields less revenue than a tax based on potential earnings because higher earners can lower their incomes to pay less taxes. But it is the best countries with relatively effective tax administrations can do.
In the “third best” scenario, the government neither has the capacity to know even actual incomes nor to collect taxes. Taxpayers may not truthfully report their incomes and tax administrations cannot collect taxes without great effort and cost. This scenario is closer to reality, especially of lower income countries like Pakistan.
Pakistan’s current tax policy operates as if it were in a “second best” world, but it better fits into the “third best” category. This has real consequences for how much tax revenue can be collected.
Take the example of corporations. Right now, Pakistan taxes corporate profits. This would be an effective way to tax in a “first” or “second best” setting because the government would know how much actual profit the corporations make and would tax them. But with the current weak system of publicly available company accounts and the inability to verify the accuracy of corporate profits, corporations have a strong incentive to under-report profits and pay fewer taxes.
If Pakistan took a “third best” approach, what would it do instead? One solution is a minimum tax on corporate revenue. Pakistan already does this for corporations whose profit is too small to tax.
With the help of collaborators, I conducted a study of the minimum tax and found that it prevented 60 to 70 per cent of the profits of the corporations in the study from going misreported. This evidence implies that a “third best” tax policy — in which the minimum tax is implemented more widely — would leverage the revenue-raising power of the minimum tax while minimising its distortionary costs to the economy. Implementing this would not entail a strenuous restructuring of the entire tax administration. This is a policy change that could be done easily with existing tax capability.
Once the government recognises that it cannot observe people’s income and trust them to report directly, it opens up more areas to apply a “third best” approach. For example, using consumption behaviour and salaries as alternative sources of information to widen the tax net would make greater sense.
The withholding tax — which is levied in Pakistan through a tax deducted at source from a payment of income — demonstrates that, to a certain extent, this is already acknowledged. It has been somewhat successful in Pakistan because the government only has to monitor employers who act as the government’s withholding agents, and employee salaries are less likely to be misreported to them.
Reforms to improve the capacity of tax administration to collect information and levy taxes will undoubtedly be important in the long term. But to support Pakistan’s development agenda now, the smartest approach could in fact be the “third best”.
Published in The Express Tribune, January 7th, 2017.