Blog
Tax, spending, and poverty

Michael Keen, Deputy Director of the IMF's Fiscal Affairs Department, will be the keynote speaker at this year’s WIDER Development Conference on Public Economics for Development. Back in 2014 we interviewed him at our offices in Helsinki, and asked him about his work on poverty, tax, and spending.

Why general subsidies are not a good way to help the poor

In the interview, Keen acknowledges that subsidies on things such as food and petroleum might seem like a tempting way to reach the poor. After all, the poor spend a larger proportion of their disposable income on these things than the rich, and it seems sensible to see these policies as pro-poor in nature.

However, Keen points out, the rich spend a larger absolute amount on food and petroleum, and most other products that might be subject to subsidies. As a result, most of the benefit in absolute terms goes to the rich, not the poor. Keen’s research shows that for every 100 dollars a government forgoes in revenue for a food subsidy, the poorest get 4 dollars of the benefit. The question for Keen then is, are there better ways to use that 100 dollars to help the poor?

Alternatives to subsidies

In advanced economies with established structures of social support the answer can be relatively straightforward. Governments can simply choose not to forgo the 100 dollars, and use it to bolster social transfers to the poorest in society.

Keen accepts that this may be more difficult to do in developing and emerging economies where social support systems may not have the capacity to reach the poor effectively. However he argues that even the very blunt tools of redistribution available in these countries are often better for reaching the poor than reduced taxes and subsidies.

Keen also suggests rethinking the common view that VAT is a regressive tax. While it may be true that VAT in itself is regressive, if the revenue raised is spent on pro-poor measures it is not necessarily the case.

Keen also highlights property tax as a good example of a tax where there are potential gains in both fairness and efficiency terms. Even if we leave aside the potential redistribution effect of a property tax it can be quite an efficient tax base, especially in developing countries where tax collection capacity is weak, because property is more difficult to hide than income.

Practical matters often act as barriers to research

Keen argues that to a large extent practical issues are the barrier preventing research on how to help the poor turn into policy. He gives the example of the UK, where a zero VAT rating of food in the UK is generally recognized as a very ineffective way of helping the poor, but it is politically untouchable.

He highlights two reasons why it is so hard for politicians to take on these policies. First, it is the rich who gain the most absolute benefit from subsidies; they are therefore a natural and powerful constituency to campaign for their continuation.

Second, it is an issue of trust in the government. While the poor may be better off if the government removed subsidies, or raised VAT, and spent the revenue on pro-poor measures, people have to believe that these other measures will actually materialize in order to support such policies. In countries where trust in government is low this can be a significant problem.

Considering both sides of the coin

The overarching message of Keen’s interview is that when thinking about how to help the poorest in society, we need to consider both the tax and spending side of the public economics coin. While a given tax may make the poor worse off, if the revenue raised is spent in a way which helps them, the overall policy package can be progressive.

The importance of considering tax and spending together is one of the key motivations behind the WIDER Development Conference on Public Economics for Development, which you can read more about here.

The views expressed in this piece are those of the author(s), and do not necessarily reflect the views of the Institute or the United Nations University, nor the programme/project donors.