DFID turns to “Payment by Results” for rural water supply in Tanzania

Turning money into water has always been the goal of water sector donors. They have money, and they want to use it to improve access to clean and safe water. Sounds simple enough.

But as my colleagues at Twaweza found in a recent policy brief, not much progress has been made on rural water supply in Tanzania, despite a massive increase in spending in the sector. Money flowed, but water only trickled.

Turning money into water? (from WaterAid, around 10 years ago)

Turning money into water? (from WaterAid, around 10 years ago)

In truth, it’s far from easy. Getting the engineering right is only one part of the challenge – managing funds correctly, avoiding corruption, deciding which communities should be prioritised, and maintaining newly water constructed waterpoints all add to the difficulty. This blog has documented many of these challenges over the past few years, particularly the struggles of the $1bn Water Sector Development Programme.

So the folks at DFID have come up with a new approach, called Payment by Results, or PbR, which they are about to start applying to rural water supply in Tanzania. And they’re backing it with big money: £150m (Tshs 400bn/-) over 5 years.

The basic idea of PbR in this case is that local government authorities (LGAs) will be paid according to the progress they make at improving access to clean and safe water. If the LGA has 100 functioning waterpoints one year, and 120 the following year, they will get extra money for the extra 20 waterpoints. How much hasn’t yet been decided, but is likely to be between £1,000 and £3,000 (Tshs 2.8m/- to Tshs 8m/-) per additional waterpoint. [UPDATE 28/7/14: DFID have confirmed the amount as £1,500 per waterpoint – approximately Tshs 3.8m/-]. This gives the LGA a clear incentive both to deliver new waterpoints and to keep older ones functioning.

Half the DFID money will be given up front as “input support”, with most of the rest held back as the PbR incentive. And payment will be made based on reports from LGAs, once these have been checked by an “independent data verification contractor”.

Part of the attraction to DFID is that they stop carrying the risk. Previously, if DFID gave money to an LGA or the Ministry of Water and something went wrong (a dry borehole, wasted money through inefficient management, etc.) the money was already spent. Under PbR, it will be up to the LGA to ensure that such problems get sorted out, or they won’t get the money.

Given the lack of past success when donors give money for rural water supply in Tanzania, it makes sense to try something new. If DFID were proposing to simply put this much money into the WSDP basket fund, I would be arguing pretty strongly against, but I think this initiative deserves cautious support.

I particularly like how it makes no distinction between installing a new waterpoint and fixing an old one. Previously, LGAs worked with contractors to install new waterpoints, then handed them over to the community and absolved themselves of all responsibility to make sure they keep working. Around a quarter of new waterpoints stopped working within two years – a major reason for the underperformance of the sector. But under this new system, keeping water flowing brings more money into the LGA. Hopefully they will give this more attention.

So what are my reservations?

Does Tshs 8m/- for the district water department really represent an incentive?

There’s an assumption that district water engineers are motivated by bringing more money into the LGA. There’s no personal incentive, no performance bonus. In my experience, some water engineers would be motivated by bringing in extra money to the department, so they can do a better job serving the community, but those are probably the same water engineers who were already working hard.

And is the amount big enough? New waterpoints can cost a lot more than proposed incentive.

The verification process will not be easy, and could easily lead to disputes

Sometimes a waterpoint that functions one day dries up or breaks down the next, and vice-versa. Who is going to adjudicate when there are differences of opinion?

Broader financial management issues could undermine the PbR incentive

The decision to target the incentive at LGAs tacitly implies that DFID and/or the Ministry see failures at that level as the main obstacle to effective delivery in the sector. That is not necessarily correct.

In the past, for example, central government has held the purse strings tightly, only releasing funds for particular centrally-approved projects, or not transferring money until very late in the financial year. If LGAs don’t have enough discretionary funding to work on installing new waterpoints and repairing old ones, they have no means to earn their “reward”.

Further, it is not clear how the results-based money will interact with other sources of funding that LGAs can spend on water supply. If other funds from central government or donors are reduced when LGAs get their reward, then the incentive is lost. Indeed, if LGAs that aren’t part of the scheme get “easier” money from other sources, then there is even a disincentive for getting involved at all. For it to work, this has got to be additional money.

And even assuming that it is additional money, if it arrives at LGA level and gets quickly swallowed up by the District Executive Director’s “need” for a new Landcruiser, (or the Director reduces the amount he allocates from other sources to water), the incentive is gone.

DFID say they will watch out for these effects, but that will not be easy. There’s a huge variation in funding streams between one year and the next, and the quality of available data on local government budgets is poor.

LGAs might try to abuse the system

Payment is based on the number of additional waterpoints, not how many additional people get water, so LGAs will be tempted to go for the low hanging fruit.

What’s to stop a District Water Engineer from installing three unneeded new standpipes in a village that’s already well served, just to get the extra money? Why spend Tshs 100m/ on drilling two new boreholes in Mahongole village (where there aren’t any waterpoints already), when you could spend Tshs 100,000/- to add two new taps to the existing piped network in nearby Idodi (where everybody already has access)? The need is greater in Mahongole, and I assume DFID would prefer that option, but the incentive is clearly to focus on Idodi.

Similarly, the incentive is to ignore more remote places or those with more difficult hydrological conditions.

If DFID try to cut out this kind of thing, it could easily become another source of disputes. If they don’t, payment by results could end up working against the interests of the poorest in society.


Despite these concerns – which DFID also recognise – I’m delighted that DFID are putting big money behind much-needed new ways of working on rural water supply in Tanzania. And I really hope that it works.

With that in mind, let me finish with two quick suggestions that could help things along.

First, DFID could publish a list of each LGA’s qualifying waterpoints, along with the calculations made for the incentive. Transparency would create opportunities for citizens, the media, civil society and politicians to scrutinise the numbers – which could keep both the LGAs and the independent verifiers honest in their reporting.

Second, keep those antennae up – by which I mean DFID should invest time and money in quality monitoring this initiative. I don’t mean mid-term or final evaluations (though these are important), so much as keeping an ear to the ground throughout the process. Watch how the programme plays out in practice – are LGAs prioritising low hanging fruit over serving the poor, are they trying to abuse the system, are they focussing more on sustainability, are the incentives effective or are they being blunted by other financial management issues, etc?

We shouldn’t kid ourselves: even with both transparency and careful monitoring, there are plenty of things that could go wrong. Expectations of PbR within DFID – particularly in London – are high. I hope they have the patience and responsiveness to work through the problems that will inevitably arise, and to make this work. Good luck to them.

One thought on “DFID turns to “Payment by Results” for rural water supply in Tanzania

  1. MJ

    Interesting stuff. I’ve always been a fan of CoD Aid, so fascinating to see the detailed challenges when you actually try to put it into action.

    However, if you follow the logic of CoD Aid it doesn’t have to matter if the District Executive Director diverts some or all of the bonus money so long as the District Water Engineer reports to the Director. Indeed arguably such ‘soft’ corruption resolves the problem about whether more money for their dept would motivate the Engineer.

    In fact I would be tempted to follow this logic further. If the bottleneck between central and local government is a big part of the problem, then why not just pay the money to the Ministry for Water, and leave them to sort out the money flow issues?

Comments are closed.