Are we killing with kindness or wise to be lending a helping hand?

What have we to show for 50 years of aid-giving by rich to poor countries? Have they resulted in any discernible improvement in the conditions of the world's poor or has the money largely gone to waste? Worse still, might aid have actually harmed these countries, resulting in a worse economic performance, as some critics have argued? What, if anything, can be done to make aid work better?

These issues are examined and discussed in these two books, both by well-known authors who are concerned with the problem of global poverty and have a considerable knowledge of the subject matter. The books' publication is timely, given the renewed interest in the subject of aid in recent years. The "Make Poverty History" campaign, the adoption of the United Nations Millennium Development Goals and the commitment of world leaders to increase aid levels at the Gleneagles Summit in 2005 have given a new prominence to the topic.

Both authors are, broadly speaking, aid optimists in that they believe that aid can play an effective role in reducing world poverty and helping the poorest countries improve their economic performance. However, both argue that a lot needs to be done to make aid work better.

The books are very different. Does Foreign Aid Really Work? is a treatise on the subject, written by an author whose earlier Foreign Aid Reconsidered was essential reading for any student of development economics two decades ago. Roger Riddell is one of the world's leading authorities on the subject. He has taught, written, published and spoken on it for more than 30 years and has directed one of the leading charities involved in aid-giving. His book contains everything anyone might want to know about the subject.

Above all, his book sets out what we do and still do not know about the impact that aid has on the countries that receive it. What emerges from a reading of the book is the sheer complexity of the topic and the great difficulty of determining whether or not giving aid has a positive impact.

First, aid is given by a large number of different donors, which includes governments, multilateral aid institutions and, of increasing importance, a multitude of different non-governmental organisations. Some of this comes in the form of aid projects, some in the form of particular programmes and the rest as technical assistance or capacity building. Second, the data required to assess aid and measure its impact often do not exist in the required form. Third, there is the difficulty of agreeing on the aims and purpose of aid and, therefore, the criteria against which performance should be measured. Fourth, there is the problem of the appropriate time horizon over which to judge the outcome. Last, but not least, there is the problem of determining what would have happened if aid had not been given - the counter-factual problem.

Perhaps not surprisingly, Riddell is sceptical over any bold claims about the effectiveness of aid one way or the other. Much research has been carried out, but the results are mixed. Perhaps more valuable than learning what works is discovering what does not work. This is to be expected as there are so many factors that affect the outcome of aid given. Therefore, Riddell wants to turn the question around, asking what can be done to make aid more effective.

Five problems need to be addressed, he argues: the amount of aid given; the rational and efficient allocation of aid; stable flows of aid; eliminating overlap, duplication and inconsistency arising from the multitude of different bodies giving aid; and addressing the imbalance in the donor-recipient relationship that leaves recipients as the junior partner. Much is already being done to address some of these issues. What matters is their cumulative impact.

By way of contrast, Making Aid Work is a short book, consisting of a series of contributions by different academics and practitioners. Like Riddell, the lead author, Abhijit Vinayak Banerjee, professor of economics at the Massachusetts Institute of Technology, is an aid optimist, who thinks that aid can be made to work.

However, what is needed is an improvement in the method of giving aid. Banerjee blames "institutional laziness" - aid institutions fail to make the effort to find out what forms of aid work before they give it. He cites the example of a recent publication of the World Bank that contained a long list of best-practice measures for reducing poverty, without providing any evidence that they work. Research to find out what does work suffers from the difficulty that we cannot know whether or not what we observe to be taking place is the result of the measures introduced. Banerjee's solution is to borrow from medical research the practice of carrying out randomised trials before giving aid and fund only projects that are proven to work.

He gives the example of randomised evaluations to find the cheapest way of getting children to spend more time in school. This turned out to be giving children deworming medicine so that they were sick less often.

Choosing the right option can matter a great deal in ensuring that scarce aid money is not wasted on expensive, ineffectual projects. As several of the other contributors show, there are practical problems in carrying this approach too far. However, Banerjee is surely right to argue for a more scientific approach to aid giving and the selection of suitable projects for funding.

Here are two books that are essential reading for anyone interested in the subject of aid and wishing to be informed about the issues involved. At a time when politicians are being urged to match actions with words by increasing levels of aid giving, it is vitally important that we stand back and ask how we can make the funds currently available achieve much better results. Making aid more effective matters as much as giving more.

Nigel Grimwade is principal lecturer and head of economics at London South Bank University.