Written By Gavin Hayman, 21 Aug 2015

Photo: Thomas “Flower Bouquet” CC BY 2.0

A huge bunch of flowers to Rick Messick for his excellent post asking two key questions about open contracting. And some luxury cars, expensive seafood and a vat or two of cognac.

Our lavish offerings all come from Slovakia, where in 2013 the Government Public Procurement Office launched a new portal publishing all its government contracts. All these items were part of the excessive government contracting uncovered by journalists, civil society and activists. In the case of the flowers, teachers investigating spending at the Department of Education uncovered florists’ bills for thousands of euros. Spending on all of these has subsequently declined: a small victory for fiscal probity.

The flowers, cars, and cognac help to answer the first of two important questions that Rick posed: Will anyone look at contracting information? In the case of Slovakia, it is clear that lowering the barriers to access information did stimulate some form of response and oversight.

The second question was equally important: “How much contracting information should be disclosed?”, especially in commercially sensitive circumstances.

These are two of key questions that we have been grappling with in our strategy at the Open Contracting Partnership. We thought that we would share our latest thinking below, in a post that is a bit longer than usual. So grab a cup of tea and have a read. We’ll be definitely looking forward to your continued thoughts on these issues.

So, how much information on public contracts should we be sharing?

Marco Siqueiros, former Head of the PPP Unit in the Brazilian state of Minas Gerais, recently boldly blogged that there is no reason not to open up all contracts and publish everything. Transparency eases public fears about contracting, allowing for a larger market for privately provided public services. In Minas Gerais, all public-private partnership contract documents are fully disclosed, along with feasibility and risk-analysis studies, projected costs, performance indicators and requirements. The move doesn’t appear to have spooked investors: indeed, the state’s strong fiscal-management system, as well as contract publication, led Standard & Poor’s in 2012 to raise the state’s credit rating to investment-grade.

Nonetheless, Rick expressed a wider concern that too much information on the government’s plans and budgets would harm its commercial interests during a bidding round. This is relatively easy to address: simply disclose who won and why they won after bidding is concluded.

Government’s need to publish some information at the upstream end of the procurement process (during the tendering round) such as the fact that there is a tender coming up and answer appropriate queries. Most of the key information including contracts, price and winning bidder should be disclosed after the bid is awarded.

The default to publication should be open

So far, so easy. Should you publish full contracts? Our answer is, that in most cases you should.

There may be some circumstances where there it is indeed not in the public interest to publish specific details of public contracts that touch on national security, privacy laws and some commercial secrets. However, these are few and far between. That is the conclusion of an excellent paper by Charles Kenny from the Centre of Global Development who analyses such exemptions in details. What we need to avoid is a kind of lazy default where information is routinely classified as confidential, especially where other checks and balances on misuse of government or corporate power are proportionately weaker.  

Data from Australia, where government contract publication is routine, are pretty revealing. Even in probably most sensitive department – the Australian Department of Defense – only 2.2% of contracts needed to be redacted.

Many governments already routinely deal with contract disclosure in Freedom of Information requests. Best practice is that the redactor should demonstrate why information needs to be removed. Their request should then be weighed by an independent party who also considers the wider public interest. Most company interests are protected by patents and copyright and, given the voluntary nature of a contract with a government, swingeing commercial protections are rarely justified. Even with redacted information, the ‘metadata’ about a contract, such as its existence or who spent public money with which companies for what outcome, is relevant to the public.

Contracts in the extractive industries

Even in the somewhat benighted oil and mining sector, where many large companies routinely argue against improved disclosure, and contracts often contain sweeping confidentiality provisions, contract publication is becoming more routine.

Over 12 countries now encourage the publication of contracts (though implementation is patchy). You can even buy contracts from websites such as the Barrows Company or download over 700 of them from this amazing open source project from OpenOil.

So far, no-one has demonstrated a credible impact on the commercial interests of the companies concerned (prove me wrong Big Oil!). More progressive companies such as Tullow Oil or Rio Tinto have recognised this shift and say that they will now push for contract publication – with government permission – themselves.

As this excellent report on the contract disclosure explains, disclosure is also useful to build legitimacy and to protect your investments. It is also a crucial step in governments learning negotiating better deals (which may be why some interests don’t like it).

Disclosing information on land deals

Speculative land deals are another example frequently cited where transparency may be harmful. The argument goes that if the government reveals the existence of a land redevelopment project, speculators may jump in and buy up the land. Personally, I am not sure that arguing in favour of the current, brutal existing information asymmetries in many land deals is really convincing. In many cases, speculators already have access to insider information, even if it is nominally secret, subverting official processes and enabling patronage and rent-seeking. This excellent investigation by Global Witness lays these problems bare.

Meanwhile, local citizens, who often own the land or have use rights, are kept in the dark until the bulldozers move in. We need robust participatory mechanisms that respect everyone’s needs and rights, not a system favouring speculation.

Open contracting involves meaningful disclosure of appropriate information at an appropriate time in the process to promote orderly and appropriate participation and consultation. We support an open source data standard that helps facilitate this and we are also teaming up with Columbia Center for Sustainable Investment on a forthcoming guidance note on better practices in contract disclosure in the land sector.

How to tackle collusion

A more substantial challenge to open contracting is that it might incentivize collusion, by facilitating monitoring of bids by cartel members.

Collusion and cartels are unfortunately all too common in both developing and developed world. A study in US markets found the two most common forms of collusive agreements were:

  1. Side payments – I will pay you to let me win etc – and/or subcontracting and joint venture arrangements from bidding firms to compensate other cartel members for refraining from bidding.
  2. Cartels that divided the market among members by selectively bidding on the basis of taking turns, territory, or market segment and/or by firms’ submitting bids known to be uncompetitive with the (cartel-selected) winning bid.

Publishing bidding information as it comes in during a procurement could aid collusion, but we aren’t asking for that (see earlier). Publishing information after decisions have been made would actually help detecting collusive arrangements. It allows you to comb the current and past data for systematic patterns. Are we seeing single source contracts that are routinely more expensive than competitive procurements? Are bids unusually low relative to previous awards? Does only one company keep winning and then subcontracting the same set of its competitors? You can see a brilliant example of analytics here from the Corruption Research Centre in Budapest. In short, it’s a valid concern but, if anything, open public data on contracting helps rather than hinders cartel detection.

Early evidence from open contracting

It’s also worth looking at some evidence from the real world. Has opening up public contracting been harmful to governments? Like in Minais Gerais, it seems the experience has been positive so far. In Slovakia, not only was money saved but frauds were uncovered and private sector competition was boosted for government contracts.

Opening up public procurement allows businesses to better track contracting opportunities and access information on what, when and where opportunities might be and award will be made. Information on past contracts also lowers the barriers to entry for smaller businesses. Seeing what previous contracts look like demystifies the procurement process and saves time and money. This is the central rationale behind the UK’s new ContractsFinder portal.

In Ukraine, our Open Contracting Data Standard has been built into the ProZorro Initiative, a multi-stakeholder public private partnership to enhance transparency and accountability of the government’s procurement system. This is really important as public procurement was one of the main vehicles for looting of the state in the past (see examples here, here and here). Early results are very positive with full disclosure and data of the contracting process, potentially large savings (10%-20%) per contract and an increase in the average number of companies bidding per procurement.

Over time, these savings mount up. Korea’s KONEPS e-procurement, which is one of the most open and transparent in the world (although it doesn’t yet publish full contracts) is estimated to be saving the government US $8 billion a year.

So we think that there is case for for opening up public contracting information and publishing most public contracts.

But will anyone use the information?

The evidence suggests that it will: by journalists, civil society organizations, businesses, citizens, and even the government.

As the Slovakian teachers have shown, citizens are interested in this kind of information when it concerns them. This is reflected in the numbers: A survey by TI-Slovakia found that one out of 10, as much as 430,000 Slovakians, had viewed a government contract or invoice online in the last year. The main contract registries get around 46,000 visits a month from citizens. Before the reforms, only 4% of the population had even filed a Freedom of Information request.

Local level innovations also suggest an appetite to use the data and for real results to flow from it.

In Afghanistan and the Philippines, civil society members have monitored hundreds of projects to deliver roads, schools and clinics. In Afghanistan, the teams were able to resolve 80 percent of the problems uncovered. In the Philippines, the price of delivering textbooks to children has halved. Indonesia Corruption Watch has developed an online platform to monitor and flag government procurement at risk of corruption and fraud. Similar initiatives exist in Uganda or Nigeria. But clearly, more can be done to support local organizations willing to dive into contracts to understand why services aren’t delivered. Sharing knowledge and supporting local initiatives is an important element of our strategy.

We know that business want access to government contracts because they are willing to pay for it. In the U.S., for example, a company called DelTek processes Freedom of Information requests for government contracts as part of an effort to help its clients win more government work. It boasts a contracts database of more than 1.7 million entries. As mentioned earlier, there are similar private databases for oil, gas and mining contracts. If firms know what previous contracts look like, it will help them bid for new work or licenses – or avoid bidding – if they can’t compete.

However, most of its use may probably come from government itself, as we have seen in the discussions with our counterparts for our Showcase projects in Mexico City and Ukraine. Professor Bob Conrad of Duke University wrote on this topic here for oil and mining contracts, especially given the way that production sharing contracts have shifted risks from the resource producing governments to resource owning governments (and thus citizens).

The balance so far has been positive. What we have seen in Slovakia, Ukraine, UK, Korea and elsewhere shows that contracts data are being used systemically. It saves government money, helps boost services, and increasing participation of businesses. You can read more on some of the early impacts of open contracting here.

With more and more data available in the future, we believe there will be lots of opportunities to put this data to work and use it to create innovative solutions that ensure more competition, better monitoring of the procurement process and the implementation, and better participation by citizen in issues they care strongly about, such as public transport, basic services, or safety.

Many thanks again to Rick for his blog: his questions were bang on.

At the Open Contracting Partnership, we’ll be looking forward to partnering across the field, busting silos and looking into these issues in-depth, as generating learning and evidence about what works will be key to switching the global default in government contracting from closed to open.

As ever, we would love to know what others think in the comments below, and you can always reach us here.