Interview with Benjamin Loevinsohn on Contracting for Health Services

a: There are a few different capacities that are needed:

(i) The procurement (or recruitment) of private sector services is a specialized skill but one that can be learned. Each financier, whether government, donor, other development partner, etc., has different rules that govern the procurement process and these need to be understood well. The contracting toolkit (available as a pdf file on this website) talks about some of the basics. My sense is that people tend to mystify procurement and make it more complicated than it should be. Don’t be intimidated!

(ii) Monitoring and evaluating private provider performance is really central to the success of contracting and requires a number of sources of information to be in place and the skilled people who can interpret the results. The sources of information include: (a) monitoring field visits; (b) HMIS; (c) health facility surveys; and (d) household surveys. These sources of information have to be used and there need to be discussions with contractors who are performing poorly.

(iii) Financial management and prompt disbursement. The payer needs to ensure prompt payment of the contractors (say NGOs) because delays can lead to gap in services. The contract management group needs to be able to understand the financial reports that contractors send them. However, it’s really important that, wherever possible, lump-sum contracts are used because otherwise contract management units can spend a huge amount of time focused on looking at inputs. Lump-sum contracts also maximize flexibility and innovation.

On the question of where these capacities should reside: There needs to be a unit that manages contracts that is properly staffed and supported. Whether this should be in the MOH or another part of the purchaser’s organization is hard to know. It does make sense to have third parties actually carrying out the health facility and household surveys to ensure independent and hopefully neutral assessment.

b: It’s difficult to measure these capacities really, but an important indicator is the promptness of payment to contractors.

c: Most countries have these skills, although they may not reside in the Ministry of Health or even in the government. They may require that (local) consultants are hired with special expertise in the three basic areas described above. There is also the opportunity to train the purchaser’s staff.

Overall, I would say that contract management is a challenging area that deserves considerable attention. However, it is not a show-stopper. Even in countries that managed contracts poorly, contractors, particularly NGOs, were still able to make progress. Also, the fact that even countries with little contracting capacity at the beginning, such as Afghanistan and Cambodia, were able to develop these skills and do a reasonable job of contract management shows that it’s possible to do this in difficult circumstances.

a: The issue of cost-effectiveness is clearly important and a little complex. However, in a number of cases, such as Pakistan and Bangladesh, NGOs performed better than governments even when given the same budget or less. For example, in Pakistan the cost (to the government) per out-patient visit was 50% less in the contracted district than in the comparison government run area. The transaction costs are difficult to quantify for both the contracting approach and government delivery of services. Then we need to look at under-the-table and indirect costs which make things complicated.

In Cambodia, cost effectiveness depends on whose costs you look at. If all you care about is the government’s own expenditures then contracting was more expensive and may not have been cost effective. However, if you also look at out of pocket expenditures by the community, contracting becomes very cost effective, i.e. better performance at lower cost to the society.

b: On the issue of equity (something critics of contracting have often brought up): There is limited data but what data exists indicates that contractors will do a better job on equity particularly if it is included in the contract as part of the terms of reference. In Cambodia for example, contracting was much more equitable than government provision of services. So to in Bangladesh and both these contracts had explicit language about equity. In Pakistan, use of services by the poor was the same in the contracted district compared to government provided services. In the Pakistan case there was no mention of equity in the contracts.

I don’t know the contents of the various contracts or agreements that African governments use with confessional NGOs or networks so I can’t comment in detail. From what I know the nature of the contracts certainly varies from country to country and has changed over time. My sense from the few that I’ve seen is that the contracts, or really agreements, focus more on resource transfers, financial accountability, and reporting, but I’m sure there are others that deal with other issues as well.

The fact that many African governments already have a long experience working with confessional NGOs suggests that there’s an opportunity to build on this experience to quickly strengthen service delivery.

Addressing quality is really important and it should be clear in the contract which standards and guidelines the contractors need to adhere to. Measuring quality, both in the private sector (whether for profit or not-for-profit) and the public sector, requires health facility assessments (see task 10 page 40 of the contracting toolkit). These are technically challenging to design but are critically important and some examples already exist.

a) There are now many examples people contracting out essential health care services. In Pakistan some 50 million people live in districts where publicly-financed primary health care services are managed by NGOs. In Afghanistan, 20 million people receive their primary health care by NGO delivered services.

b) In the contracting toolkit you’ll find some an example of a good contract on page 107 and some examples of terms of reference that you could include in a contract. You can use the checklist in the toolkit (on page 67) to look at existing contracts and see how they might be strengthened.

a) There are a number of governments, often with donor support, that finance P4P schemes. The World Bank is working with governments in 8 countries in Africa to test P4P.

b) It’s possible that it would help strengthen health service delivery. It’s an approach that should certainly be tested on a reasonable scale to see how well it works in the Kenyan context.

Sustainability issues are always difficult to discuss and usually complex. The first issue regarding sustainability is whether the intervention or approach is worth sustaining, i.e. does it work? Is it more effective than other approaches? Sustaining failure is un-wise.

The next issue is whether the intervention is cost-effective than the alternatives but even here it is complex because if you’re reaching more remote communities or poorer people, you may be willing to support an intervention even if the costs per patient served are higher.

Sustainability often has to do with the resources that are available which is really a political question about the priorities of the government.

All told I don’t think we understand very much about the sustainability of P4P yet. Let’s first make sure that it works well in different contexts!!

I know that in Rwanda, the quality of care was explicitly part of the P4P approach and I agree with you that it is critically important. It’s not helpful if providers focus on quantity of services and cut corners on quality.

It's important to be clear about geographical responsibility of the private providers. If you use a competitive bidding process then you can assign responsibility. A completely different way of doing things is to just work with all existing private providers and pay on the basis of services provided. This admittedly is complicated.

This is an important question. Again, geographical responsibility needs to be clear. However, there are different models of contracting, such as contracting in of management services that can be used in existing public systems. For a more detailed discussion of this issue you should see pages 12-15 of the contracting toolkit available on the website.

This is similar to a previous question. If the purchaser is clear about which geographical areas need to be covered, then the contracts can be be clear about the areas that need to be covered. In Africa, confessional NGOs may not want to cover areas where they are not currently working, but that can be overcome through negotiations and clear terms of reference.

I think so far the evidence indicates that contracting can improve the quantity of care that is delivered, even in very difficult circumstances. There is some evidence that contracting can imporve quality of care and also equity, i.e. that it is better at reaching poor people. This is especially true if the contracts specifically address issues of equity and quality of care in addition to the quantity of services provided.

You may want to look at the answer I provided to Tricia Moser. Also see pages 52-55 of the contracting toolkit for more specific suggestions. Clearly there's a need for a diverse set of skills and it's more than just a part time job!!

These are lot of good questions:
a) I think the commonest mistakes in contract management are (i) not monitoring contractor performance adequately and (ii) not paying contractors in a timely fashion. However, I've seen a number of instances where these functions have been handled very well. For example, in Afghanistan the contract management unit really whipped some poor performing NGOs into shape and they dramatically improved their performance. This was done through field visits and extensive discussions with NGO managers.

b)Monitoring of contracts requires careful technical assessment. Sometimes it makes sense to use a third party expert.

c) If you mean contracting to individual private providers, there are many examples in OECD countries. There are fewer in developing countries but the work done on TB case management by private providers is potentially interesting.

d) Contracting should be considerred where public providers are not performing optimally. This may be particularly true for remote and under-served rural areas and poor urban ones.

a) There are actually a lot of examples in Africa, but unfortunately they have not often been written up and evaluated rigorously. There's a published article about nutrition services being contracted in Senegal (see page 156 of the contracting toolkit).

b) That's tough to answer but an important question. There's relatively little experience with contracting with for-profit providers. This is an area we will have to learn more about. Some cuation may be required because health service contracts, by their nature are not complete. On the other hand many OECD coutries use implicit contracts with for-profit providers all the time.

I think there are many advantages to having formal agreements. Of critical importance is: (i) agreeing on what the objectives are and how progress will be measured; (ii) the roles and responsibilities of both parties; and (iii) clear mechanisms for resolving disputes or differences. My sense is that formal contracts simply work better because both parties take them more seriously.

I think that the word "contract" sometimes scares people off because it implies a commercial arrangement with for-prfit entities. I'm not sure that label really matters. What's important is that the agreement, contract or MOU contain the elements that are needed to achieve results. This is what the contracting toolkit is all about. Using the checklist on pages 66 and 67 will give a quick sense of whether the MOU, agreement, contract is addressing all the important issues.

Below is the answer I provided to a previous question about cost-effectiveness. On the obstacles, I think there are many: (i) lack of familiarity with contracting, people are scared of what they don't know so well; (ii) some people in governments see contracting as an implicit insult, i.e., that they're doing a good job so why involve NGOs or others. Often this is a question of pointing out that private providers often have more flexibility and so can respond more easily becuase they don't face civil service rules and regulations; and (iii) some unscupulous people in governments see contracting as loosing power, prestige, or the opportunity to steal money. For a fuller treatment of this issue see pages 21-24 of the contracting toolkit.

The issue of cost-effectiveness is clearly important and a little complex. However, in a number of cases, such as Pakistan and Bangladesh, NGOs performed better than governments even when given the same budget or less. For example, in Pakistan the cost (to the government) per out-patient visit was 50% less in the contracted district than in the comparison government run area. The transaction costs are difficult to quantify for both the contracting approach and government delivery of services. Then we need to look at under-the-table and indirect costs which make things complicated.

In Cambodia, cost effectiveness depends on whose costs you look at. If all you care about is the government’s own expenditures then contracting was more expensive and may not have been cost effective. However, if you also look at out of pocket expenditures by the community, contracting becomes very cost effective, i.e. better performance at lower cost to the society.

Thanks Aneesa, this was lots of fun!!