Government contracts regarding the use of public property and finances should be published by default. Many jurisdictions already require that contracts be made public in response to requests for the information; some now publish contracts proactively. Doing so helps new entrants compete in the market for public contracts, helps governments model their projects on other successful examples, and allows citizens greater insight into how their taxes are being spent. This provides a practical outline for reaping the benefits of open contracts while addressing legitimate concerns about costs, collusion, privacy, commercial secrecy, and national security.
Related: Brief | Blog | Working Group
Working Group on Contract Publication
Akilagpa Sawyerr, Laryea, Laryea & Co., P.C.
Alan Hudson, Global Integrity
Caroline Nicholas, United Nations Commission on International Trade Law
Charles Kenny, Center for Global Development
Chrik Poortman, Construction Sector Transparency Initiative
Chris Taggart, Open Corporates
Gabriel Sipos, Transparency International
Jeff Gutman, Brookings Institute
Joe Powell, Open Government Partnership
Lesley Coldham, Tullow Oil
Mima Nedelcovych, Initiative for Global Development
Marcos Siquera-Moraes, Minas Gerais Public-Private Partnership Unit
Mathias Huter, Independent
Michael Jarvis, World Bank
Patrick Heller, Revenue Watch Institute
Tina Soreide, Chr. Michelsen Institute
William Savedoff, Center for Global Development
Members of the working group were invited to join in a personal capacity and on a voluntary basis. This paper of the working group reflects a broad consensus among the members listed above, but their membership should not be taken as endorsement of every opinion or recommendation in this report. This paper does not necessarily represent the organizations with which the working group members are affiliated, the Center for Global Development, its funders, or its board of directors.
Thanks to Nancy Birdsall, Frank Anthony Fariello, Krzysztof Izdebski, Richard Messick, Tam Nguyen, and Ousmane Deme for valuable input.
The Center for Global Development is grateful to the Omidyar Network and the UK Department for International Development for support of this work.
It is time that the world’s citizens learn whether they are getting what they paid for from government contracts.
For rich and poor countries alike, getting government contracting right is an economic necessity. For countries in the Organisation for Economic Co-operation and Development, government procurement of goods and services is worth around 15 percent of GDP; the share is similar in many lower-income countries. And contracts with oil, gas, mining, and agricultural firms to exploit state-owned natural resources frequently involve billions of dollars in a sector that produces profits equal to 5 percent of global GDP. Beyond its sheer financial scale, government contracting involves provision of public services and goods that are central to the development process.
But government contracting often goes badly—and sometimes particularly badly in poorer, smaller countries. On the procurement side, poor price forecasting, limited competition, poorly designed tenders, and (too often) fraud and corruption combine to produce low-quality, overpriced, and over-budget delivery. A notorious case was the Dabhol Power Plant, built in India’s state of Maharashtra by a consortium led by US energy firm Enron. It produced power at a cost up to four times that charged by local producers under a power purchase agreement that forced the local utility to buy it. The annual cost to the state was estimated at $1.3 billion. 
On the licensing side, limited knowledge of the market, poorly designed social and environmental conditions, and (sometimes) malfeasance produce agreements that have high external costs and low benefits to the public. Even deals negotiated with probity and efficiency can appear suspect if they are kept secret. In 2014, a leaked contract between Norway’s Statoil and the Tanzania Petroleum Development Corporation covering an agreement to share natural gas production suggested more generous terms to the company than had been signaled by “model agreements” previously made public. The leak created considerable political turmoil in the country even though outside experts suggest the deal may not be out of line with international precedent.  Because most of those precedent agreements remain confidential, it is difficult for Tanzanian civil society to evaluate that claim.
Transparency can help improve the quality of government contracting. It lowers barriers to entry for firms to bid on work by providing information about previous similar contracts and gives bidders greater comfort that the bidding process will be fair. Governments benefit from increased competition as well as the experience of similar contract models from other jurisdictions, which should improve price or revenue forecasts as well as the quality of contract specifications. Civil society can use contract information to ensure that delivery of services (or the supply of revenues) matches the agreement.
This report documents examples of the benefits of contract transparency: a 50 percent increase in competition for government tenders in Slovakia, reduced variation and lower average prices in hospital supplies in Latin America, lower costs for social housing in France, the exposure of significant political party funding by sole-source contract winners in Georgia, and civil-society monitoring of a social development fund by a mining company in the Democratic Republic of the Congo.
In particular, contract publication by all countries can be of particular help to small, poor economies. Total government expenditure in Liberia is $815 million per year (a little less than the cost of one medium-sized power plant). Perhaps one-third of that involves contracted services. Experience with large contracts is rare, and being able to draw on an international stock of similar contracts would be of particular value. In negotiations with global multinational firms over billion-dollar deals including extractives concessions, power purchase agreements, and large infrastructure projects, contract transparency will help level the playing field. The World Bank’s Road Costs Knowledge System takes data on the prices for construction and rehabilitation of roads from World Bank–financed transport projects and allows users to calculate the average cost of materials and work across countries by type of project: for example, repaving a two-lane highway over flat ground when the cost of cement is $110 a ton. Such information allows for better cost estimation, improved budgeting, and a stronger negotiating position with potential suppliers.
Small, poor economies also have less capacity to monitor contracts, so they benefit disproportionately from third-party monitoring by both local and international civil society. Those that are large aid recipients see a considerable portion of their service delivery and (even) infrastructure provision contracted by foreign governments and agencies. It is only right that the supposed beneficiaries of those contracts and their government representatives are able to see what is being contracted on their behalf.
This report lays out the growing consensus behind proactive contract transparency. International agreements including the UN Convention against Corruption require transparency throughout the contracting process. Many countries allow the text of contracts to be reactively published in response to petitions similar to a Freedom of Information Act (FOIA) request in the United States. Governments in countries including Georgia, Slovakia, the United Kingdom, and Colombia are proactively publishing the entire text of all contracts soon after signature. And more countries proactively publish contracts in particular sectors involving public-private partnerships and extractive industries.
But there are concerns regarding proactive contract publication. In particular, these involve the risk that publication might assist collusive bidders, the protection of commercial and national secrets that might be revealed by publication, and privacy issues.
The Center for Global Development Working Group on Contract Publication brought together representatives from civil society and the public and private sectors to discuss these concerns. The working group concluded that there was little evidence that publication would support collusion. And while there were legitimate issues around secrecy and privacy, most contracts would not raise them. For the minority of contracts involved, valid secrecy and privacy concerns could be addressed using a redaction and classification regime that would be a comparatively minor burden on both bidding firms and government officials. The case is laid out in detail in this report.
A number of developing countries have led the way in contract transparency and demonstrated that proactive publication is both feasible and desirable. Donor countries should catch up—both by publishing their own contracts and by providing support to other developing countries to follow suit.
Center for Global Development
Summary and Recommendations of the Working Group
Government procurement worldwide is worth around $9.5 trillion a year. Oil, gas, and mining rents (the gap between the price of the goods produced and the cost of production) amount to around $5 trillion, which is 4.8 percent of global GDP. Governments routinely sign multibillion-dollar contracts regarding the use of public property including those natural resources. The resulting contracts are public documents, for which default practice should be in favor of publication.
There are concerns, however, about the costs of the publication process, the risk that publication would support collusion, and the benefits of a stand-alone contract publication initiative. Commercial, national-security, and privacy concerns also exist. This report of the Center for Global Development Working Group on Government Contract Publication provides guidance about contract publication and, in particular, recommendations around secrecy and security issues.
Benefits of Publication
Publication of the contract text itself is part of a broader open-contracting agenda to increase transparency and monitoring capacity throughout the selection, operational, and review phases of contracting. Many countries and sub-national jurisdictions already require by law that the content of contracts be published in response to a Freedom of Information request. Some countries now publish contracts proactively, including the federal governments of Colombia, the United Kingdom, Slovakia, and Georgia. This small additional step in favor of transparency could have significant benefits: for companies, new entrants can have a far better idea of the goods and services they will bid to provide if they have easy access to existing contracts; the increased quality and extent of competition fostered by contract publication would benefit governments, which could also model new tenders on successful contracts from elsewhere; and contract publication would allow beneficiaries and civil society to monitor value for money and service delivery. Evidence from Slovakia suggests procurement transparency reforms in the country that included contract publication were associated with an increase in competition on the average government tender: from 2.3 bids in 2009 to 3.6 bids in 2013.
Evidence suggests that commercial concerns and those about costs, collusion, privacy, and national security deserve consideration, but they can be addressed without undue burden:
- Costs of publication are apparently low, focused on the minority of contracts where a redaction process is likely to be necessary. Australia’s federal contract database flags contracts if they have specific confidentiality clauses. Of the contracts in the database for 2012, only 2.2 percent were flagged with such clauses. But costs are reduced and benefits considerably increased if contract transparency is part of a broader procurement transparency and civil-society capacity-building initiative.
- Collusion involves monitoring the bids of competitors, and increased transparency might theoretically make such monitoring easier, but fears that transparency would aid collusion are concentrated on the upstream end of the procurement process (during the tendering round) rather than the downstream end (after the bid is awarded), when contracts would be published.
- While there are legitimate commercial, national-security, and privacy concerns, they involve a small minority of contracts and can be addressed using a principles-based redaction policy.
The list of potential commercial interests that might be legitimately confidential in some circumstances includes novel designs and technologies, financial information, and strategic plans, among certain other information. Such information should not be present in the majority of contracts. As part of the tendering process, when submitting their bids, suppliers should be given the opportunity to identify which pieces of information in a contract they regard as being commercially sensitive and would not want published and the reasons why that is the case. Government departments should assess the information that the winning supplier has identified as being sensitive against the exemptions set out by the Freedom of Information Act (FOIA) or other supporting regulation and redact information where that is in the public interest.
Even if a number of contracts simply will not appear in a contract database under a national-security exclusion, many defense contracts could be (at least partially) published, and information should be excluded using existing rules for classification.
Existing experience with contract publication suggests a number of legitimate privacy concerns, in particular regarding information on third parties that might receive services under a contract but are not signatories to it. Again, this affects a minority of contracts and should involve only partial (principles-based) redaction.
The Conclusions of the Working Group
As part of the broad consensus in favor of government transparency, publishing government contracts has a number of potential benefits that justify its limited cost. These include improved design, tendering and price forecasting, increased quality and extent of competition, and improved monitoring of value for money and service delivery.
Proactive publication is better than reactive publication because it levels the playing field among bidders, maximizes the information in the public domain, and reduces the transaction costs involved in obtaining (and providing) information under FOIA or similar requests.
The presumption should be in favor of publishing all contracts, including licenses, concessions, permits, and grants; any other document exchanging public goods, assets, or resources (including all annexes, schedules, and documents incorporated by reference); and any amendments or side agreements thereto at all levels of government. This will help level the playing field among competitors, maximize the information in the public domain, and avoid displacement of information from contracts to side agreements.
The presumption should be in favor of timely (within days of signature), current, and routine publication of full contracts, with exceptions or limitations narrowly defined by law. This will increase the information in the public domain at a time of its maximum utility and reduce burdens on decision-makers regarding what to publish.
There should be clear rules about timing and incentives to publish, including a specific deadline after signature. This will reduce burdens on decision-makers, aid compliance, and ensure a level playing field among firms.
Rules regarding publication, redaction, and exclusion should be principles based as opposed to mechanical and exhaustive. The level of (legitimate) concern regarding unredacted publication is likely to vary between contracting and procurement models and the goods and services being contracted. That legitimate concerns are likely to be so different suggests the advantage of a principles-based approach. In countries with Freedom of Information legislation, the principles should follow naturally from the exemptions laid out in that law.
It should be clearly visible which purchases or which aspects of a contract are not disclosed and why and which contracts have not been published and why except in rare cases involving national security. This will aid compliance and the appearance of compliance regarding redaction procedures.
Commercial confidentiality is a valid concern for a small part of a small number of contracts, and countries should establish appropriate procedures that delineate and allow redaction for those instances. Given the existence of patent and copyright protections and the voluntary nature of participation in a contract, broader protections are rarely justified. While there may be contracts that contain commercial secrets that are in the public interest to maintain, that public interest should be clearly demonstrated and minimal redaction should apply.
There are legitimate privacy and national-security concerns with full publication that should be managed, in particular regarding information about third parties that might receive services under a contract but are not signatories to it. Similarly, there are technologies, approaches, and other information that might be described in contracts touching on national security that it is in the public interest to redact. But these concerns involve only sections of a small proportion of all contracts.
The onus should be on firms to delineate which sections of a contract they consider commercial in confidence, and governments should review that request on the basis of Freedom of Information legislation (which should favor maximum disclosure). Commercial secrecy is the primary concern of contractors, not the government, so firms should suggest what information they believe might be commercially sensitive. The role of the government should be to review the redaction request through the lens of the public interest.
Governments have the primary responsibility to ensure that national-security and privacy concerns are addressed in the redaction process. As the public body in the contracting relationship, it is clear that government should take the lead in addressing these public-good issues. Confidentiality on the grounds of national security should operate using standard classification regimes (which will create a legal liability for the release of such information).
The disclosable nature of the contract and related documents should be clarified within the contract itself. This will ensure all parties are aware of responsibilities and the standing of contracts.
There should be a clear dispute-resolution mechanism. If there is a dispute about disclosure, the case should be rapidly reviewed by an independent information commissioner who has security clearance or another specified impartial body with the capacity to respond rapidly (potentially the court system), preferably following standard Freedom of Information procedures. This will increase both the efficacy of and trust in the publication regime.
Documents should be open (without copyright) and machine readable, while metadata (including price, date, contracting parties, descriptions of goods and services provided) should be developed and data should be published in a user-friendly format. This will maximize the utility of information contained in contracts to all stakeholders. There is no tradeoff between publishing the full text of contracts and metadata about those contracts—the activities are complementary.
Contract publication should not stand alone, and countries should follow broader open-contracting standards. Publication will and should be pursued as part of a broader open-contracting agenda covering planning through procurement and execution, as well as budget, disbursement, and corporate transparency, and involving capacity building for legislators and civil society. Without these additional steps, the utility of information released by contract publication will be significantly diminished.
As part of the effort to increase contract transparency, countries should make commitments to publication as part of Open Government Partnership National Action Plans and negotiate for greater disclosure through the United Nations Commission on International Trade Law (UNCITRAL) and the World Trade Organization Government Procurement Agreement, building on the principle of transparency enshrined in these documents. Suitable private and public-private organizations including the World Economic Forum, chambers of commerce, and the UN Global Compact could help advance the agenda by creating registries of companies declaring support for contract publication as part of a nondiscriminatory system with suitable protections—preferably alongside governments committing to contract transparency. Citizens pay for government contracts. It is time they knew what they are paying for.
Government procurement worldwide is worth somewhere in the region of $9.5 trillion a year. Publishing the underlying government contracts related to this procurement has a number of potential benefits: for companies, new entrants can have a far better idea of the goods and services they will bid to provide if they have access to existing contracts; the increased quality and extent of competition fostered by contract publication would benefit governments, which could also model new tenders on successful contracts from elsewhere; and contract publication would allow third parties in civil society to monitor value for money and service delivery.
In addition to procurement contracts, publishing lower-value purchases as well as sales of property and grant and subsidiary contracts alongside contracts providing rights to exploit public lands and oil, gas, and mining rights will have similar beneficial effects. Oil, gas, and mining rents (the gap between the price of the goods produced and the cost of production) amount to 4.8 percent of global GDP alone—or around $5 trillion dollars.  Most of these rents are distributed on the basis of contracts between governments and extractive industries.
Government contracts are public documents, where the presumption should be in favor of publication. For this reason, there is a widespread global norm in favor of transparency in procurement, licensing, and concessions. The UN Convention against Corruption, signed by a considerable majority of the world’s countries, requires transparency throughout the procurement process, and other international regimes have an equivalent requirement.  The UNCITRAL model law requires all terms of contract that are known at the bidding stage to be disclosed to bidders and the award of the contract and its price to be published with very limited exceptions. This suggests the considerable majority of information in the considerable majority of contracts is already seen as being rightfully in the public domain in many countries. 
Again, the principle that the contract is a public document that can be published is already enshrined in law in many cases around the world. In the United States, for example, contracts can be—and regularly are—reactively published as the result of a request under the FOIA. Some governments now proactively publish contracts, including the federal governments of Colombia, the United Kingdom, Slovakia, and Georgia. , At the local level, the Australian state governments of New South Wales and Victoria and the US county government of Miami-Dade, Florida, also publish contracts. Other jurisdictions are considering a similar change, suggesting the start of a trend.
The proposal discussed here—to publish the full text of contracts—is a logical next step in the widely supported global agenda of procurement transparency. And it is part of a broader open-contracting effort to increase transparency and monitoring capacity throughout the planning, budgeting, procurement, operational, and review phases of contracting.  Guidelines are likely to be helpful along a number of dimensions including accessibility and metadata standards. But perhaps most immediately required are guidelines about what should not be published—exceptions to a general rule of proactive publication covering issues including privacy, commercial secrets, and national security.
The existing cases of proactive contract publication suggest that such exemption guidelines are frequently (and unsurprisingly) based on Freedom of Information guidance. But existing experience also suggests that Freedom of Information guidance itself is often applied inconsistently across FOIA requests for contracts—experience in the United States, for example, suggests some of those charged with redaction see all price information in a contract (even for standardized commodities with a public price) as a commercial secret while others see little or no information about prices or other contract details as excludable under Freedom of Information exceptions. The purpose of this report of the Center for Global Development Working Group on Government Contract Publication is to provide guidance about contract publication and, in particular, recommendations around secrecy and security issues with publication in an attempt to delineate a broad consensus around what is possible to publish and why exemptions should be limited.
 There are exceptions. For example, Austria has no Freedom of Information law. Combined with a data protection law that protects companies alongside individuals, this limits the government’s ability to publish limited information such as contract prices.
Status of Publication Regimes Worldwide
More than 100 countries worldwide have Right to Information or Freedom of Information laws.  Many of those Freedom of Information regimes already allow for reactive publication of contracts. For example, competitors for government contracts in the United States regularly request previous versions of that contract to help prepare bids,  and winning bidders are advised to mark parts of the contract they might consider commercially sensitive in the event of a FOIA request. A brief review of the texts of 13 Freedom of Information Laws from countries that do not proactively publish contracts  suggests that while there are sometimes provisions made for the protection of commercial secrecy or the bid process in such laws,  contracts are in no way explicitly excluded from the general provisions of the act.
Some countries have gone further toward proactive publication:
In countries where routine publication is still not the norm, some sectors are moving ahead on the proactive publication agenda. The (official) standard in Latin American countries reviewed by the World Bank Institute is for routine and complete proactive public-private contract publication without redaction.  De facto practice does not appear to follow this rule, although many contracts are indeed published redaction-free. In the Brazilian state of Minas Gerias, all Public-Private Partnership (PPP) contract documents are proactively fully disclosed, and in most cases, the original scanned document is downloadable. In addition, the PPP website presents a project summary, technical background papers and feasibility studies, projected costs, results from demand studies, detailed information about risk analysis, performance indicators and requirements, guarantees, and requirements for project financing, among others. The full versions of monthly performance reports are also available alongside a comparative chart of an “aggregated performance index” indicating how well the service provider is delivering services. 
In extractives, countries such as Democratic Republic of the Congo, Guinea, Sierra Leone, and Mozambique have taken significant steps toward large-scale disclosure of contracts, and Mexico and Niger have enshrined disclosure of extractive contracts in their constitutions. Additional disclosing countries now include Afghanistan, Azerbaijan, Liberia, Peru, Republic of Congo, Timor-Leste, Niger, Sao Tomé and Principe, Senegal, Mali, Burkina Faso, Cameroun, Mauritania, Madagascar, and Peru.  Companies including Tullow Oil and Kosmos, as well as several mining companies, have become proactive in stating their willingness to publish contracts where governments are agreeable. For example, the government of Ghana asked Tullow to publish production agreements for the Jubilee project, and these can be found on the company’s website.
There appears to be movement toward breaking the circle of blame operating between many governments and private-sector operators: governments often point at the protection of commercial interests as the main reason for confidentiality. But firms in the same country report they are positive toward transparency. For example, the United Kingdom’s Confederation of British Industry has called for greater transparency. The group’s Public Service Strategy Board declared, “All government contracts should be published online, as long as the customer is happy for this to happen. And where commercial information is redacted, there should be a clear explanation of who requested the redaction and why.” 
Meanwhile, language coming out of the World Bank Group, regional multilateral banks, and the International Monetary Fund, alongside the inclusion of an “encouragement” of disclosure in the Extractive Industries Transparency Initiative standard, suggests a growing international consensus behind proactive publication.
It is worth noting that while existing contract publication systems are not perfect, with evidence of missing contracts and (sometimes) considerable redaction, they do demonstrate the institutional feasibility of contract publication. Furthermore, Freedom of Information requests routinely lead to the release of whole contracts, so the proposal to publish proactively is only a change in process for the majority of countries with FOIAs rather than (necessarily) in what is considered publishable information.
 Australia, Canada, Croatia, Czech Republic, Estonia, Germany, Guinea, India, Indonesia, New Zealand, Nigeria, Poland, and South Africa.
 In the case of Nigeria, information that might “frustrate the bid process.” See the following for laws: http://www.right2info.org/laws#section-19.
 See Sistema Electronico para la Contratacion Publica (SECOP; 2012). The law requires contract publication for contracts larger than 50 times the minimum wage, made by any of the public institutions defined in an older law: “The Nation, regions, departments, provinces, the Capital District and the special districts, metropolitan areas, municipal associations, indigenous territories, and municipalities; public facilities, industrial and trading enterprises, societies mixed economy in which the state has more than fifty percent (50%) participation and indirect decentralized entities and other legal entities in which there is such majority public participation, regardless of the name they adopted, in all areas and levels.” See http://www.alcaldiabogota.gov.co/sisjur/normas/Norma1.jsp?i=28776. All contracting activity can be found using SECOP’s search tool: https://www.contratos.gov.co/consultas/inicioConsulta.do.
 For example, all the relevant information for some construction services contracted by the National Agency of infrastructure are available online, including amount; contractor; relevant dates; a bunch of technical, financial, and legal documents required to bid; and the contract as such, with no redaction whatsoever. See https://www.contratos.gov.co/consultas/detalleProceso.do?numConstancia=12-1-85398.
 Alongside local authorities and “other parts of the wider public sector,” intelligence agencies, financial and nonfinancial public corporations, parliamentary bodies, and devolved administrations are all exempt. Note there was a complaint about the Slovak publication system that it was not initially clear enough as to which entities had the obligation to publish contracts. See http://www.legal500.com/c/slovakia/developments/21611 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/61200/guidance-publication-of-new-central-government-contracts.pdf.
 According to Alexander Kashumov of the Access to Information Programme in an e-mail to the FOIANet Listserv, May 28, 2014.
 http://politicsofpoverty.oxfamamerica.org/2014/08/absolutely-kenya-president-backs-full-oil-contract-disclosure/ .
 CBI (2014).
Why Make Publication Routine?
As suggested by the opening language of most Freedom of Information legislation worldwide, the default for information generated by governments is that it should be public. Governments act in the name of and for the benefit of citizens; as a result citizens have a right to know what the government is doing on their behalf. At the same time, there are exceptions to this principle, and the benefit of publishing government contracts should outweigh any harms. Potential benefits accrue to firms, government, and civil society. Potential risks include time and expense, easing collusion, and limited impact thanks to partial disclosure. These are discussed in the next sections, although it should be noted that the evidence base on the impact of contract publication remains limited.
For the private sector, evidence of the value of past contracts to firms considering bids is clear from the numerous Freedom of Information requests for contracts made by firms in the United states, and a company dedicated to processing such requests as part of an effort to help its clients wins more government contracting work.  Similarly, there exist pay-access databases regarding revenues and other features of oil, gas, and mining contracts. 
Being able to review past contracts issued for similar work (or for a similar concession) allows firms to bid on new contracts only where they have a realistic chance of winning and to put in higher-quality bids at lower cost. As bid preparation costs range from about 0.5 to 1.0 percent of contract values on large construction contracts (and often considerably higher than that on smaller consulting contracts), this can be a significant benefit. 
The importance of better information to encouraging firm participation is illustrated by an analysis of highway procurement auctions in Oklahoma, which studied a change in policy toward releasing the state’s cost estimate for the work to be undertaken prior to bid submission. The policy change led to lower average bids and a lower winning bid. The effect was larger in more complex and uncertain projects.  The policy allowed new entrants to put in more realistic bids and increased their survival rate in the industry, suggesting it reduced the information advantage of incumbent firms regarding likely costs of work. 
By building trust in the probity of the procurement, licensing, and concessions system, a high level of transparency in contracting may help reassure firms that their bids will be evaluated on a level playing field. Firms also may benefit from reduced uncertainty as a result of contract publication. They form the basis for a fact-based discussion around terms and conditions with parliamentary and civil-society groups while reducing the risk of future governments’ renegotiating or reneging on secret deals of their predecessors. In turn, this may lower the cost of capital. And transparency may ease public fears around contracting out, allowing for a larger market for privately provided public services. 
For government, by making the bid and execution process less uncertain and less costly, contract transparency should increase firm interest in bidding on government tenders. This increases competition and should reduce prices and improve the quality of bids.  Analysis by Transparency International (TI) in Slovakia suggests that procurement transparency reforms in the country that included contract publication were associated with an increase in competition on the average government tender: from 2.3 bids in 2009 to 3.6 bids in 2013. It also suggested that greater competition is linked to lower prices for goods, a finding repeated elsewhere. 
There is further direct evidence that publishing contract price data can lead to reduced average costs and reduced cost variance in procurement. The city government of Buenos Aires, Argentina, began publishing data on prices paid for a range of common medical supplies purchased by hospitals in 1996. Prepublication prices varied by a factor of 10 for ethyl alcohol and needles, for example.  After the announcement that prices would be published, the average price paid dropped—as did the variation in prices paid, which approximately halved. 
Easy availability of past contracts will also allow for better design (utilizing successful past contracts as models) and more accurate planning and budgeting. In India, for example, nearly half of all roads projects see cost overruns greater than 25 percent,  and the average roads project in a worldwide sample saw cost overruns of 20 percent.  The World Bank’s Road Cost Knowledge System (ROCKS)  provides data on unit costs road works including resurfacing, upgrading, and widening based on data culled from World Bank–financed contracts. The aim of ROCKS is to improve the reliability of new cost estimates and reduce the risks generated by cost overruns. With routine contract publication, databases like ROCKS could be replicated on a far larger scale at the country and global level.
Transparency may be a particularly important tool for government contracting in cases of noncompetitive and nonstandard contract award, in providing assurance to all parties that the contracting process was fair.  If it increases trust in the procurement system, contract transparency allows increased flexibility—reducing red tape to allow smaller firms to bid or using procurement designs that allow for negotiated procurement. Given that the standard least-cost award based on detailed terms of reference or tender documents is often an inefficient method of contract award, this flexibility can also improve procurement outcomes. 
The European Union recently revised its procurement rules to allow negotiated procedures (where agencies publish their requirements, accept bids, and then negotiate with multiple bidders toward a revised bid, finally awarding the contract to the bidder who, postnegotiation, proposes the best solution). Given the greater discretion inevitable in such a negotiation process, negotiated procurements are legally required to provide additional transparency in terms of publication of selection criteria and greater information provided to losing bidders about why they were not selected. Analysis of social housing contracts from Paris suggests that negotiated contracts combined with greater transparency reduced bid prices by 26 percent and also reduced the probability of bid renegotiation. 
With regard to citizens and civil society, contracting material can be of considerable interest to the general public. Colombia’s contract site had nearly half a million visitors a month in 2008.  In Slovakia, a recent survey by TI suggested that as much as 9 percent of the population had viewed a government contract or invoice online in the past year. TI Slovakia staff suggest that contract publication also has made civil-society monitoring procurement far more straightforward—not least, removing the time-consuming (two- to three-year) process of Freedom of Information court verdicts in cases wherein these are necessary prior to information release. 
TI Slovakia was able to use data culled from the contract publication system to uncover considerable inefficiencies in hospital procurement—including the purchase of identical CT scanners in the same year for prices that varied by as much as 100 percent. Among other results, the analysis found that more than half of all hospital tenders attracted only one competitor, competitive (as compared to single-bid) tenders were associated with prices that were 20 percent lower and lower final cost variance,  there was considerable variation between hospitals in the level of competition for their tenders, and electronic procurement was associated with considerably increased competition. 
In Georgia, TI analyzed 430,000 government purchases that were single sourced and awarded without a tender and cross-referenced the data with the company registry, asset declarations of public officials, and party donations records. The investigation found at least US$150 million in single-sourced purchases going to companies owned by members of parliament and public officials or their spouses and found that in 2012, 60 percent of donations disclosed by the ruling party came from owners, directors, and lawyers of companies that had received contracts without tenders. TI Georgia suggested the average donation was worth about 4 percent of the contract values involved. 
In extractives, journalists, companies, and community representatives in countries from Peru to the Democratic Republic of the Congo to Ecuador have begun to take advantage of publicly available contracts to educate citizens in areas affected by projects and to make better use of contract provisions on consultation and community development projects. Information in extractives contracts of interest to citizens includes how much governments are likely to receive, tax and other incentives provided, commitments made in terms of public financing, and social and environmental provisions.
For example, in the Democratic Republic of the Congo, civil society is monitoring contract implementation of the Tenke Funkurume copper and cobalt mine including the requirement to pay 0.3 percent of net mining proceeds into a social development fund to benefit local communities. The monitoring organizations found that the company complied with this obligation from 2009 to 2012 and that the funds had been spent on tangible projects—schools, bridges, wells—that have benefited the mine’s neighbors. ,
Publishing contract details can also help civil society hold government to account for delivery. Regarding school textbook delivery in the Philippines, information about what was to be delivered where and when was used by civil-society groups including scout groups to monitor on-time delivery of the right quality and quantity of books.  Again in the Philippines, the Concerned Citizens of Abra for Good Government used project and payroll information provided by the Department of Public Works to compare progress toward project completion as suggested by payment schedules to progress and quality of provision on the ground across 600 infrastructure projects. As a result of their monitoring, 11 district engineers were found guilty of negligence, and a number of projects were completed that otherwise would have languished. Slum dwellers in the K-East Ward in Mumbai have used design information obtained through a Freedom of Information request to discover that the pipe being installed to bring water to their neighborhood was the wrong diameter. 
 Strassmann and Wells (1988).
 Note the average price result was not statistically significant.
 De Silva et al. (2008), De Silva, Kosmopoulou, and Lamarche (2009).
 This is part of the logic behind the UK Confederation of British Industry’s support for publication.
 Analysis of World Bank infrastructure contracts in Africa suggests strong relationship between low competition in bidding and cost overruns on contract delivery. Only half of the analyzed contracts saw an adequate level of competition to reduce the risk of overruns (Africon 2008).
 Savedoff (2008).
 The UK National Health Service is proposing a similar effort across hospital trusts in Britain (Green 2014). Note that the price dispersion may not be predominantly due to corruption. A study of procurement in Italy looked at generic goods purchased across a range of government departments and levels. The average price paid for goods among public bodies at the 90th percentile is 55 percent greater than the price paid by bodies at the 10th percentile. The authors estimate that “passive waste”—inefficiency—accounts for 83 percent of total waste compared to “active waste” related to corruption, which accounts for only 17 percent (Bandiera, Prat, and Valletti 2009). Note also greater price transparency on the Internet has not always had the expected impact of reducing prices and price dispersion. Evidence suggests that prices for generic goods (books, CDs) are frequently more dispersed online (even if average prices are somewhat lower), but for more complex goods (insurance, cars) prices are less dispersed (Pan, Ratchford, and Shankar 2004). Again, greater Internet penetration reduces average airline prices but not interfirm price dispersion on a given route (Orlov 2011).
 Kenny (2010).
 In 2005, $145 billion of US federal contracts were issued noncompetitively—nearly 40 percent of total federal contracting (United States House of Representatives, Committee on Government Reform 2006).
 The combination of minimum criteria on technical solutions and separate comparison of prices, for example, is a disincentive for innovation (since innovative solutions will go unrewarded). With more details released it becomes easier to defend acquisition of solutions that are superior to the minimum criteria. (Although note a potential tradeoff: the more innovative the solution, the more firms might consider it a commercial secret and resist disclosure.)
 Chever and Moore (2013).
 See SECOP (2012).
 This compared to cost estimates—that is, more competition was associated with fewer cost overruns.
 Transparency International (2013).
 Majeed (2013a).
 Majeed (2013b).
 Such misdelivery is common: a 2010 audit of 18 Zambian roads projects jointly financed by the government and donors found that substandard cement had been supplied in all projects and that in half the projects the concrete was weaker than required (International Bank for Reconstruction and Development/World Bank 2011).
Costs and Risks
For firms, beyond potential fears regarding commercial secrets to be discussed in later sections, there is a concern that selective publication might lead to a competitive disadvantage, giving competing entities access to information about how rival companies have structured previous bids/contracts. This fear should be reduced under routine publication regimes.
Firms also may face costs associated with a redaction process—marking up contracts, appealing publication decisions—as well as costs related to public engagement over contract terms. These costs may be priced into bids but apparently have not been a significant issue in current regimes given evidence of lower prices under proactive publication regimes.
For government, beyond national-security concerns discussed in later sections, there may be fears around cost, reduced competition, and the risk of collusion.
The costs of the physical publication system itself appear to be minor: the Georgian e-procurement platform, which includes contract publication, was developed in house for less than US$1 million. It serves as the document management and procurement-planning platform for all state entities alongside every school, municipality, and ministry in the country. Costs are reduced the more the existing procurement system is online—something that has been found to independently increase efficiency in a range of countries. 
The cost of time involved in the redaction process may be more significant. At the same time, under Freedom of Information regimes, government officials have to undergo this process on an ad hoc basis, and under current publication regimes it does not appear to have been a major administrative burden. The more standardized the rules (and the more circumscribed the exceptions to publication), the lower should be the administrative costs of publication.
Metadata from the federal procurement system in Australia can suggest some orders of magnitude about the number of contracts where redaction will be required. Australian government guidance about confidentiality throughout the procurement cycle states that once a contract has been awarded the terms of the contract are not confidential, unless the agency has determined and identified in the contract that specific information is to be kept confidential in accordance with the guidance about specific clauses.  Australia’s contract database flags contracts if they have these specific confidentiality clauses and provides the reason for them: costing and profit information; intellectual property; Privacy Act considerations; “public interest”; secrecy provisions; and “other.” Looking at the 74,859 Australian federal contracts in the database for 2012, with a combined value of a little over AU$40 billion, 1,676 contracts were flagged with confidentiality clauses. That is about 2.2 percent of the total. The contracts with confidential clauses were worth 6.6 percent of total contract value. If Australian experience (including both Freedom of Information and contracting practices) broadly translates to other countries, this suggests that the considerable majority of contracts should be publishable without redaction.
A final potential cost is that procurement choices may come under increased scrutiny and involve time-consuming response to citizens groups or short-term political costs. The information contained in contracts can be used by political opponents. As one example, the recently leaked production-sharing agreement between Norway’s Statoil and the Tanzania Petroleum Development Corporation governing revenues from natural gas drilling suggested a lower share to government than had been expected based on a model agreement. This caused the government considerable public embarrassment that might have been avoided if the contract neither had been officially published nor had been leaked. While this may be a real concern, it does not appear a legitimate argument for confidentiality. In existing publication regimes, the additional workload on individual procurement officials appears minimal. And in the Tanzania case, the government has chosen to avoid future embarrassment by proactively publishing agreements. In the long term, the benefits of transparency on trust and the quality of government contracts should outweigh any short-term discomfort that it might cause.
Regarding the quality of contracting, publication of contracts may lock in expectations regarding future contracts including approach, price, terms, and conditions. This may reduce contract evolution toward lower-cost, higher-efficacy approaches and more favorable terms and conditions for government. The validity of this fear is difficult to empirically verify but appears at odds with evidence presented above that transparency has as a rule reduced prices for procured items. Given contract publication is associated with greater competition and that it also should act to illuminate failed approaches alongside new approaches as well as ease replication of best practice, the overall impact may be to reduce lock-in.
There may be concerns that publication (even with redaction for commercially secret information) would deter some contractors from bidding, thereby reducing competition. Note evidence from Slovakia suggests the opposite impact of increased transparency, and there are reasons to believe publication should increase the quality and number of bids. Note also that it has not limited competition for US government employment that pay scales are published—indeed the actual pay and benefits of named employees is often a matter of public record.  Again, despite the fact that over 12,000 contracts subject to successful Freedom of Information requests are already available in a searchable database online in the United States for only $99, contractors also appear willing to bid for federal, state, and local contracts in the country.
A second concern around competition is that publication will assist efforts at overt or tacit collusion between bidders. Collusive systems rely on information regarding cartel member behavior to continue. Contract publication may make the flow of such information more straightforward. The Organisation for Economic Co-operation and Development’s Guidelines for Fighting Bid Rigging in Public Procurement illustrates this concern. It states a central aim to “design the tender process to reduce communication amongst bidders.” It proposes e-bidding, eschewing bid conferences, and “wherever possible under the legal requirements governing the award notices,” to “keep the terms and conditions of each firm’s bid confidential”—even after bids have closed in the bid opening process—as this can “facilitate the formation of bid-rigging schemes, going forward.” This suggests a potential tradeoff between transparency and collusion.
The utility of contract information published after award on maintaining a cartel is limited, forever. A study of collusion in US markets found the two most common forms of collusive agreements were (1) side payments (or subcontracting and joint venture arrangements) from bidding firms to compensate other cartel members for refraining from bidding and (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.  The concern in both cases is that bidder awareness of other potential bidders eases the formation of cartels. This can be avoided without (permanent) loss of transparency: if bid documents are available to all for anonymous and free collection or download, if procurers avoid bid conferences and nonanonymized communication with potential bidders, and if bids are submitted electronically or otherwise impersonally (by post), neither firms nor procurement agents need know who else might bid until after bid closing.
In either cartel model above, the collapse of the cartel would likely be apparent to the cartel-selected winner by the time of bid award—in that it would not win the bid. Tacit collusion can involve price signaling during the auction process—but again, whether the collusion worked would be obvious to involved parties by the time of bid award, before contract publication. 
But procurement is often a repeated game—the same companies can potentially bid on multiple similar government contracts over time. Publishing information postprocurement might theoretically allow for better monitoring of adherence to cartel rules. In practice, however, it would be impossible (and impractical) to hide who had won the bid—and the central aim of cartel bidding is to control who wins. Given that, transparency after bid opening would expose the breakdown of a collusive arrangement where bidders reneged on the agreement by either submitting a losing bid that was not designed to lose in the particular way agreed among the cartel or submitting a winning bid that was not designed to win in the way that it did. While such schemes can be imagined (e.g., involving floor prices for particular elements of the bid, which the winning contractor subsequently renegotiates before signature), the working group was unaware of an actual uncovered collusive scheme where this was a major factor.  Note also, in any case in which the procuring agent is involved in cartel formation, he or she will have access to all information regardless. And in any regime with the right to access contracts under Freedom of Information legislation, cartels could ensure compliance through FOIA requests. 
Furthermore, publication provides a new tool for government competition watchdogs who can red flag procurements and look for indications of collusion in metadata. Georgia’s e-procurement system highlights procurements that are found to be high risk by algorithms and that then can be scrutinized by experts at the procurement agency. A large enough database of contracts regarding similar projects could help uncover contract price inflation linked to cartel activity or corruption. 
Ex post contract publication may create some theoretical tradeoff between monitoring adherence to cartel rules and external monitoring of cartel formation, then. But especially in environments where the procurement agency itself is potentially involved in forming and monitoring cartels the benefits of ex post transparency on competition and reduced collusion appear to outweigh the risk. The burden of proof should lie with those officials who believe contract transparency in a particular sector or bid process will considerably increase the risk of collusion to make that case.
It is also worth noting that a number of recommendations by the Organisation for Economic Co-operation and Development on bid rigging would reduce the burden of contract publication—not least, designing tender specifications on the basis of functional performance rather than a specific design (which should reduce commercial secrecy concerns in the contract itself) and electronic procurement systems (which will ease publication).
For legislators, citizens, and civil society, beyond privacy concerns discussed later, there is the fear that (absent consistent and full publication, broader release of information around procurement and performance, and capacity building for those monitoring procurement) contract publication will have little impact. It would be a fig-leaf reform that did not affect accountability or outcomes.
With regard to selective publication (of previous regime contracts or arbitrarily selected contracts), consultations in developing the open-contracting principles produced a consensus in favor of “timely, current and routine publication” of contracts with “any exceptions or limitations narrowly defined by law.”
But for all of the benefits of publication, it should be clear that contract transparency will be no panacea for corruption or poor contracting outcomes. Colombia’s contract publication regime did not prevent a military procurement scandal that cost the commander of the armed forces his job in 2014, for example. This speaks to the importance of nesting publication in a broader open-contracting agenda of procurement transparency, civil-society strengthening, and broader government probity.
The concern regarding the value of transparent contracts’ standing alone without metadata and other supporting documents and information suggests that alongside contracts, basic information and data about public investment plans and spending, project design and development objectives, tender documents, payment, and delivery of contracted goods and services should all be easily available (and easily linked to particular contracts).
With regard to capacity, competitors, scrutiny officers, and financial analysts will all be able to use published contracts as monitoring tools, but legislative bodies and civil society more broadly may need additional capacity building. This suggests the value of models such as the Extractive Industries Transparency Initiative and the Construction Sector Transparency Initiative that match increased information release with support for civil society to use that information.
Both benefits and full costs of contract publication are hard to estimate with any degree of accuracy and will likely vary considerably by contract type, sector, and country. At the same time, the above analysis suggests the benefits are likely to outweigh comparatively small costs and risks. Given the widespread presumption in favor of publishing government information, this suggests the basis for considerably wider rollout of contract publication accompanied by careful evaluation of impact.
 One example is the pay and benefits received by officials in Ontario’s government: http://www.fin.gov.on.ca/en/publications/salarydisclosure/2011/ministries11.html.
 Note, however, that in the absence of collusion more transparent processes are likely to lead to more efficient outcomes, suggesting that the level of transparency in the bid process should vary according to the risk of collusion (Hendricks, McAfee, and Williams forthcoming). Given the greatest risk of collusion is in concentrated markets, one response would be to limit procurement transparency in such markets. At the same time, these are also the markets where informal information flow is most likely, suggesting procurement opacity may prevent only third parties from monitoring collusion rather than the collusive players themselves. Again, concentrated markets also may be at higher risk of corruption, making transparency in such markets particularly valuable. Other approaches than (downstream) opacity are likely to be preferable, then, including third-party analysis of bidding patterns.
 Pesendorfer (2000).
 Bajari and Yeo (2009).
 Imagine a scheme in which bidders agree to a floor price for a particular element of the bid (gravel in a road contract, as it might be) while competing on other price elements of the bid. This would ensure the winning bidder would garner greater profits than otherwise. Publishing price details would allow cartel members to ensure the winning bidder had not won by breaking the floor price. At the same time, all bidders in a round having submitted the same (high) price for a particular component of the bid would be a fairly obvious red flag of collusion.
 Given how many countries allow contract release under a Freedom of Information Act (FOIA), one might expect that were contracts a common tool of cartel enforcement a case would have come to light by now.
 Cartel activity has been found to raise prices for roads projects by 8 percent in Florida and as much as 40 percent across a sample of 29 developing countries (International Bank for Reconstruction and Development/World Bank 2011).
General Principles of Publication
To maximize the benefits of publication and help reduce costs, certain general principles apply:
The presumption should be in favor of publishing all government contracts, including licenses; concessions; permits; grants or any other document exchanging public goods, assets, or resources (including all annexes, schedules, and documents incorporated by reference); and any amendments thereto, at all levels of government.  This presumption would include state-owned enterprises where possible, especially where these enterprises are spending public money or contracting over public resources, where the government owns a majority share of the enterprise, and where the enterprise operates as a monopoly. 
This presumption may include some subcontracts: potentially those specifically mentioned in the main contract and signed primarily to help deliver that contract and with a significant material impact on the performance of that contract.
The presumption should be in favor of timely, current, and routine publication of full contracts with exceptions or limitations narrowly defined by law. Contract summaries should complement, not replace, full contract publication.  It should be clearly visible which purchases or which aspects of a contract are not disclosed and for which reason as well as which contracts have not been published, with reasons.
There should be clear rules about timing and incentives to publish. The United Kingdom sets an advisory limit that contracts are published within 20 days of the end of the standstill period, that is, after a set period of time has passed from which the contract award decision is made until the time at which the contract is signed. Slovak legislation provides considerable incentive to ensure timely publication: a contract becomes valid only once it is published online.
Rules regarding redaction and exclusion should be principles based. The level of (legitimate) concern regarding unredacted publication is likely to vary between contracting and procurement models and the goods and services being contracted: national security, individual (as opposed to firm) contracts, software and other nonrival products, cost-plus contracts, or contracting forms where considerable negotiation is possible between any draft laid out in the terms of reference and final signature. In part because legitimate redaction concerns are likely to be so different, this suggests the advantage of a principles-based approach to redaction (as opposed to a mechanical, exhaustive set of regulations) overseen by an independent (security-cleared) arbiter. In countries with Freedom of Information legislation, the principles should follow naturally from the exemptions laid out in that law.
There should be a clear dispute-resolution mechanism. If there is a dispute about disclosure, the case should be rapidly reviewed by an independent information commissioner who has security clearance or another specified impartial body with the capacity to respond rapidly (potentially the court system), preferably following standard Freedom of Information procedures. There would be two separate tracks—the first regarding disagreements between contracting firms and government officials on commercial secrecy within the contract prior to signature (which should be rapid) and the second regarding the legitimacy of the redactions open to all parties postsignature. Models for this can be found in Slovenia and Croatia (where the commissioner recently ruled in favor of publication for contracts between state hospitals and pharmaceutical companies). An information commissioner would be subject to higher court review.
The disclosable nature of the contract and related documents should be clarified within the contract itself. Examples include the Model Mining Agreement of the International Bar Association and Denmark’s Model Hydrocarbons License. The Model Mining Agreement states, “30.1 (a) This Agreement . . . is a public document, and shall be open to free inspection by members of the public at the appropriate State office.” The United Kingdom provides guidance about model language to insert into the contract:
The parties acknowledge that, except for any information which is exempt from disclosure in accordance with the provisions of the FOIA, the content of this Contract is not Confidential Information. The Client shall be responsible for determining in its absolute discretion whether any of the content of the Contract is exempt from disclosure in accordance with the provisions of the FOIA. Notwithstanding any other term of this Contract, the Contractor hereby gives his consent for the Client to publish the Contract in its entirety, (but with any information which is exempt from disclosure in accordance with the provisions of the FOIA redacted) including from time to time agreed changes to the contract, to the general public. 
Documents should be open, machine readable, and accompanied by metadata. The United Kingdom specifically mandates that documents are made available under an open government license in an open, nonproprietary, standardized format to ease reuse by third parties.  There is no tradeoff between publishing the full text of contracts and metadata about those contracts—the activities are complementary. Metadata should be developed (cost, contracting parties, procurement codes defining what is being purchased, flags for redactions, and so on), and data should be published in a user-friendly format, preferably through an application programming interface.
Contract publication should not stand alone. Publication should be pursued as part of a broader open-contracting agenda covering planning through procurement and execution and involving capacity building for legislators and civil society.
 With regard to the United Nations Commission on International Trade Law (UNCITRAL) Model Law, the Guide to Enactment states,
The enacting State may extend the application of the Model Law to certain entities or enterprises that are not considered part of the government, if it has an interest in requiring those entities to conduct procurement in accordance with the Model Law. In deciding which, if any, entities to cover, the enacting State may consider factors such as the following: (a) Whether the Government provides substantial public funds to the entity, or a guarantee or other security to secure payment by the entity in connection with its procurement contract, or otherwise supports the obligations of the procuring entity under the contract; (b) Whether the entity is managed or controlled by the Government or whether the Government participates in the management or control of the entity; (c) Whether the Government grants to the entity an exclusive license, monopoly or quasi-monopoly for the sale of the goods that the entity sells or the services that it provides; (d) Whether the entity is accountable to the Government or to the public treasury in respect of the profitability of the entity; (e) Whether an international agreement or other international obligation of the State applies to procurement engaged in by the entity; (f) Whether the entity has been created by special legislative action in order to perform activities in the furtherance of a legally-mandated public purpose, and whether the public law applicable to government contracts applies to procurement contracts entered into by the entity.
 Whole contract publication removes discretion from officials in deciding what is material information with regard to the contract. Whole publication also allows for easy replication of the contract by future bidders or other jurisdictions and detailed examination by third parties as to whether contracts were fully and efficiently delivered.
Areas for Redaction
It is expected that in the majority of contracts there is no need for any redaction at all. For example, in South America, where there is limited room to negotiate contracts during procurement, a draft of the signed contract would have been disclosed already as an annex to the Request for Proposals. The only information that is added to the contract in the standard case is the price. The same broad system applies to standard procurements in many other parts of the world—and the universal norm is moving toward freely published tender documents.
There will be exceptions to this general rule, however—for example, where the contractor has proposed and the client accepted innovative materials or techniques that appear in drawings or specifications and where there are security or privacy concerns. In the Australian Federal Contracts database, nearly one-half of the contracts with confidentiality flags listed costing and profit information as the reason for the flag. Ten percent raised intellectual property concerns. Privacy Act and public interest considerations accounted for another 13 percent between them. Statutory secrets were involved in 4 percent of contracts, leaving “other” with a 24 percent share. This suggests the majority of contracts that may (arguably) need redacting do so for reasons surrounding commercial secrecy,  but national security and privacy are other legitimate issues.
There is a diversity of views regarding the overall legitimacy of commercial secrecy concerns in government contracting, in part shaped by broader attitudes toward market functioning and the role of secrecy in competition between firms. Legitimacy of concern is likely to vary by sector and contract type: sectors that produce nonrival goods (pharmaceuticals, software, consulting) might be the most sensitive regarding price information. Service contracts including legal, financial, and general consultant services also may provoke concern because of privacy issues. Firms may be concerned about the release of cost information through bills of quantity or other elements of the contract that they see as a commercially secret element of a superior business model.
Freedom of Information regimes typically allow commercial secrecy exceptions including in the United States and United Kingdom and under the UNCITRAL model law. UNCITRAL allows the nondisclosure of commercially sensitive information that could impede fair competition under a current contract or a future contract and information arising in one contractual relationship that might affect competition in other contractual relationships.  But commercial secrecy concerns with regard to public contracts are less valid than those between private firms because of the public interest in government transparency, and the commercial secrecy issue remains irrelevant or small in most contracts.
The Australian Council of Auditors-General, in discussing contract transparency, laid out the public interest case:
Those in the private sector who wish to gain commercial advantage from dealings with the Government cannot seek to escape the level of scrutiny that prevails in the public sector. Such scrutiny is required because of the non-commercial nature of much Government activity, the non-voluntary relationship between individuals and their Government, and the different rule of law which applies in the public sector compared to the private sector. 
The Australia New South Wales Auditor General, as reported by the Victoria Parliamentary Committee, also made the case that postaward information is unlikely to be commercially privileged:
There would be a very clear demarcation between commercial information which is ex ante, before a decision is made relevant to that information, and commercial information which is ex post—that is, after decisions have been made. Tender documents provided before the tender decision is made are a particularly commercially sensitive . . . because the benefits and rights attaching to that information can be usurped by others should that information be given out. After the decision is made . . . the information is of very little value in a commercially confidential sense. 
As a practical matter, for large contracts, notes the Victoria Committee, the final provisions are known by hundreds of lawyers, advisers, financial consultants, and others—it is hard to imagine a competitor determined to know what was in the contract would have too much trouble finding out. The public accounts committee concluded that the insistence on confidentiality of clauses in contracts most frequently originated from the government, suggesting the commercial secret concern may be exaggerated. 
The UK Parliament’s public accounts committee chair, Margaret Hodge, suggested similar sentiments about contracting in Britain in 2014.  The Public Accounts Committee 2014 report, Contracting out Public Services to the Private Sector, concluded that government departments should not “routinely use commercial confidentiality as a reason for withholding information about contracts with private providers.” 
The list of potential commercial interests that might be legitimately confidential, from the UK FOIA Exemptions Guidance, section 43, is as follows:
Research and plans relating to a potential new product
Product manufacturing cost information
Product sales forecast information
Strategic business plans, including for example, plans to enter, develop or withdraw from a product or geographical market sector
Marketing plans, to promote a new or existing product
Information relating to the preparation of a competitive bid
Information about the financial and business viability of a company
Information provided to a public authority in respect of an application for a license or as a requirement of a license condition or under a regulatory regime
In standard contracts, little of this information should be present. Under good practice models, the inclusion of considerable design and process information in procurement contracts should be the exception (or at least should apply to only particular contract forms). Regarding the provision of goods or many services, the contract should preferably define outputs to be delivered rather than the inputs to be used in delivery. In those cases as well, commercial secret concerns are unlikely to arise regarding the contract. In cases where the government has good reason to define inputs in the contract (specific technologies to match legacy systems or equipment, e.g.), the related technical specifications will be part of the bid documents and so be public before contract signature. This leaves only the question of the pricing of individual items as a potential commercial secret that might sometimes be claimed by firms.
There are also methods that the private sector can use to limit the commercial sensitivity of information in contracts. Regarding commercial secrets around design and process, firms have recourse to protect intellectual property including the patent system. Or firms can create specific legal entities for work with government. For example, in the case of infrastructure public-private partnerships where concessions and licenses are frequently published, the private provider is typically a special purpose vehicle—a legal structure created specifically for the purpose of fulfilling the contract. Its cost structure, as well as other commercially sensitive information, is specific to that particular contract and project.
With regard to extractives, model contracts contain little if any information that would be considered commercially sensitive.  A review by Revenue Watch of extractives contracts concludes that references to trade secrets or future transactions are legitimate competitive harm concerns but rarely found in extractives contracts while information that is commonly found in such contracts—covering work obligations, local content rules, employment and training requirements, financial terms and payment rates, and parties to the contract—are unlikely to cause competitive harm. 
Because contracts are difficult to keep completely confidential, because they usually contain little information with regard to design and process information or costs (as opposed to prices), because of other methods to protect commercial secrets, and (perhaps above all) because of public interest in contract transparency, the commercial secret exception should be carefully delineated. While there may be contracts that contain commercial secrets that it is in the public interest to maintain, (1) that public interest should be clearly demonstrated, and (2) minimal redaction should apply. On the process to be followed regarding commercial secrecy exceptions, UK government guidance about contract publication suggests a workable model:
As part of the tendering process, when submitting their bids, suppliers should be given the opportunity to identify which pieces of information they regard as being sensitive and would not want published and the reasons why they would not want the particular pieces of information published. Departments should not use this information when evaluating bids. Once departments have evaluated the bids and awarded the contract to the winning bidder, departments should assess the information that the winning supplier has identified as being sensitive (along with the rest of the contract) against the exemptions set out by the Freedom of Information Act when considering which contractual information should/should not be published. Departments should engage with the chosen supplier in this process and inform them of the outcome of their assessment in terms of which information is to be published and which, if any, will be redacted.
The general principle should be that firms suggest which information they view as commercially sensitive at the time of bid submission and the government official should determine which of that information (and only that information) that appears in the contract it is in the public interest to keep confidential under a commercial secrets exemption. There should be a fast and independent appeals process if firms are dissatisfied with the initial determination, with the ability to walk away from the contract without penalty if the firm is still unsatisfied. 
National Security and Privacy
In the Australian federal contracts database, there are 31,871 contracts signed by the Department of Defense or the Defense Materiel Organization that began in 2012. Collectively they are worth AU$18.2 billion. Of these contracts, 869 (2.7 percent of all defense contracts) are marked with a contract confidentiality flag. They are worth AU$1.4 billion (7.8 percent of all defense contract values). This suggests that Department of Defense and Defense Materiel contracts do raise secrecy concerns, but only somewhat more often than government contracts in general.  Even if a number of contracts simply do not appear in the database under a national-security exclusion, many defense contracts could clearly be (at least partially) published.
There are technologies, approaches, and other information that might be described in contracts touching on national security that it is in the public interest to redact. In addition to defense and intelligence contracts, major infrastructure contracts might also sometimes be seen to be of national-security concern. FOIAs sometimes mandate or allow that information officers not disclose information that might prejudice the security of buildings, structures, and systems (including IT systems or infrastructure).  But these concerns involve a small proportion of all contracts and only sections of those contracts.
At the same time, there is no list under UK guidance for contract types or components that might be excluded under national-security grounds; the relevant guidance section simply suggests the FOIA exemption applies where withholding the information is “required for the purposes of safeguarding national security.” Similarly, Freedom of Information legislation tends to be vague around definitions of national-security exemptions. For example, South Africa’s Promotion of Access to Information Act suggests, “The information officer of a public body may refuse a request for access to a record of the body if its disclosure (a) could reasonably be expected to cause prejudice to (i) the defense of the Republic; (ii) the security of the Republic; or (iii) . . . the international relations of the Republic.” It then goes on to give a list of examples of such information “without limiting the generality” of the preceding language. 
The UNCITRAL Model Law and Guidance on Public Procurement provides exceptions around disclosure that would impede law enforcement, and
non-disclosure of such information as is necessary for the protection of essential security interests of the State, which may be legally identified as classified information. . . . Essential security interests may concern not only the national defense of a State but any other sectors and issues, for example security related to public health and welfare. . . . Non-disclosure might be justified by the sensitive nature of the subject matter of the procurement or by the existence of classified information even if the subject matter itself is not sensitive (for example, when the need arises to ensure confidentiality of information about a delivery schedule or the location of delivery), or both. 
It has proven difficult to develop a comprehensive list of types of information that might be reasonably redacted on national-security grounds, in part because the secrecy of such information depends considerably on circumstance. 
As with commercial secrecy, security concerns should not be considered carte blanche for redaction. Canada’s Access to Information Act, for example, suggests information release should continue if “the public interest in disclosure clearly outweighs in importance any financial loss or gain to a third party, any prejudice to the security of its structures, networks or systems, any prejudice to its competitive position or any interference with its contractual or other negotiations.”  The default minimum should be that if contract information is to be redacted on national-security grounds, that information should be formally classified using the usual procedures.
Regarding privacy, under the UK Data Protection Act and relevant FOIA guidance, it appears salary information can be released, although it has frequently been redacted in practice.  Existing experience with contract publication suggests a number of (more) legitimate privacy concerns, in particular regarding information about third parties that might receive services under a contract but are not signatories to it. The UNCITRAL Model Law provides exceptions around personal information that a public authority may occasionally be prohibited by law from disclosing—for example, public health and welfare information, depending on the nature of the market concerned, or information for which the law requires prior judicial authorization prior to disclosure. 
Many countries, including those in the European Union, have specific privacy legislation such as a Data Protection Act. Guidelines about which personal information should be redacted or excluded would follow that legislation as well as what (if anything) is valid in relevant Freedom of Information legislation. Given the restrictiveness of many data protection laws, that might suggest very little personal information in a contract could be published without specific authorization. But contract publication regimes might chose to involve standard language providing consent by named service providers or company officials to the release of personal information contained in the contract.
 New South Wales, Australia, which publishes some government contracts, applies the following rules regarding nondisclosure: commercial-in-confidence provisions of a contract, details of any unsuccessful tender, any matter that could reasonably be expected to affect public safety or security, and any information where there is an overriding public interest against disclosure. It is not clear why details of an unsuccessful tender would be included in a contract.
 For New South Wales, Australia, commercial-in-confidence provisions of a contract refer to information that would reveal the contractor’s financing arrangements, financial modeling, cost structure, or profit margins; could place the contractor at a substantial commercial disadvantage in present or future dealings with the agency; or would disclose any intellectual property in which the contractor has an interest.
 Parliament of Victoria (2000).
 Parliament of Victoria (2000).
 “All too often it’s the government that hides behind commercial confidentiality because they don’t want the service to be analysed too closely” (Green 2014).
 Bulgaria’s Access to Information Law provides specific language on when “commercial secrecy” should not apply:
Production or commercial secret may not be any facts, information, decisions and data related to business activities whose keeping as secret is in the interest of the claimants but there is overriding public interest in its disclosure. Until the contrary is proven, there is overriding public interest in the disclosure when the information: а) gives opportunity to the citizens to form their own opinion and to take part in ongoing discussions; b) improves/facilitates the transparency and accountability of bodies under Art. 3, sub-art. 1 with regard to the decisions they make; c) guarantees the lawful and purposeful fulfillment of the legal obligations of bodies under Art. 3; d) reveals corruption and abuse of power, poor management of state or municipal property, or other unlawful or unpurposeful actions or lack of actions of administrative bodies or responsible officials within the respective administrations by which state or public interests, rights or legal interests of other persons are affected; e) disproves disseminated unauthentic information which concerns significant public interests; f) is related to the parties, subcontractors, the subject, the price, the rights and obligations, conditions, terms, and sanctions specified in contracts where one of the contracting parties is an obliged body under Art. 3.
 US FOIA exemption 4 covers “trade secrets and commercial or financial information obtained from a person [who is] privileged or confidential.” Department of Justice guidance on FOIA suggests a trade secret is “a secret, commercially valuable plan, formula, process, or device that is used for the making, preparing, compounding, or processing of trade commodities and that can be said to be the end product of either innovation or substantial effort.” See http://www.justice.gov/sites/default/files/oip/legacy/2014/07/23/exemption4_0.pdf.
 Regarding price data in particular, now superseded US Department of Justice guidance about prices in FOIA’d contracts suggested that
the prices in government contracts should not be secret. Government contracts are ‘public contracts,’ and the taxpayers have a right to know—with very few exceptions—what the government has agreed to buy and at what prices. … The law and policy calling for disclosure of prices in awarded government contracts is parallel to the long-established rule for disclosure of government employees’ salaries. See http://www.justice.gov/oip/blog/foia-update-disclosure-prices. There is a different standard in dealings with the government. In the private sector, disclosure of salaries or prices is usually at the option of the parties. But the common principle—where the government is a party—is that neither individual privacy nor private commercial interest justifies secrecy as to government commitments of public funds.
Subsequent case law in the United States has made the position less clear cut, and firms have successfully reverse-FOIA’d so that line-item prices in some circumstances would count as causing competitive harm to a company. Nevertheless, companies would need to demonstrate that potential harm (revelation of profit margins, e.g.) to justify redaction of such information from a FOIA’d contract. See http://www.justice.gov/oip/blog/foia-post-2005-treatment-unit-prices-after-mcdonnell-douglas-v-air-force.
 The one exception to the general rule that firms should propose and government should accept or reject proposed redactions may involve collusion if a case can be made that contract publication will support it. But in cases where a contracting official believes there is a strong case that release of contract information in a particular contract or set of contracts would aid collusion, they should have to make that case to the independent arbiter in charge of reviewing publication decisions.
 Note these numbers may not be strictly comparable with those reported earlier as the data analysis occurred a year later based on updated data from the contracts database. Note that only 26 contracts raised a confidentiality flag under the basis of statutory secrecy, but potential concerns with national security appear to have frequently been covered by other confidentiality flags (e.g., a contract covering forward-looking infrared software upgrades for F/A-18 A/B aircraft was flagged “Intellectual Property”). At the same time a number of contracts covering weapons orders were not flagged confidential, perhaps because the contract did not disclose confidential information about their capabilities.
 This is the case with Uganda’s Access to Information Act, for example. See http://www.right2info.org/laws#section-19. See also Canada’s Access to Information Act: “The head of a government institution may refuse to disclose any record requested under this Act that contains information that could reasonably be expected to facilitate the commission of an offence, including, without restricting the generality of the foregoing, any such information … the vulnerability of particular buildings or other structures or systems, including computer or communication systems, or methods employed to protect such buildings or other structures or systems.” http://laws-lois.justice.gc.ca/eng/acts/A-1/page-8.html?texthighlight=security#s-16.
 http://ico.org.uk/for_organisations/guidance_index/~/media/documents/library/Freedom_of_Information/Detaile d_specialist_guides/safeguarding_national_security_section_24_foi.ashx.
 The examples relate to the quantity, characteristics, capabilities, vulnerabilities, or deployment of weapons, weapons designs, military units or personnel, or intelligence capabilities. See http://www.right2info.org/laws#section-19.
 UNCITRAL (2012).
 For example, fuel purchases might not usually be considered confidential, but if the US Department of Defense were purchasing large quantities of jet fuel at a base near a particular zone of geostrategic tension, it might not want to publicize the fact.
 In rare cases, the release of contract information not covered by a national secrecy concern might be reasonably expected to harm an individual’s safety.
Routine contract publication is achievable at a reasonable cost and can provide considerable benefits to contractors, governments, and civil society alike. The major concerns around contracting—including the risk of reduced competition as well as secrecy and privacy concerns—can usually be addressed with redaction of specific parts of the contract. Most contracts can be published without redaction, often as a comparatively small (if valuable) step toward additional transparency over existing global and national norms around publishing procurement documents.
The net benefit of publication will be increased in regimes where there is a strong presumption in favor of full publication, utilizing narrowly prescribed principles for redacting information related to commercial secrecy, personal privacy, and national security. And contract publication should be nested in a broader open-contracting agenda.
There is an international dimension to contract transparency: countries will benefit from the experience and information contained in contracts from elsewhere, providing international best practices as well as costing information.
Countries could make commitments to publication as part of Open Government Partnership National Action Plans and negotiate for greater disclosure through UNCITRAL and the World Trade Organization Government Procurement Agreement. The Conference of States Parties to the UN Convention against Corruption could adopt contract publication as an anticorruption tool and monitor implementation through peer review.
Suitable private-sector and public-private organizations including the World Economic Forum, chambers of commerce, and the UN Global Compact could help create registries of companies declaring support for contract publication as part of a nondiscriminatory system with suitable protections—preferably alongside governments committing to the agenda. Existing initiatives including the Extractive Industries Transparency Initiative, the Construction Sector Transparency Initiative, and Open Contracting provide the logical home for increased support to governments and civil society for contract publication—there is no need for a new institution to support this effort. The Open Contracting Partnership in particular could provide a useful forum for sharing experiences and further developing standards as well as collating evidence of the impact of transparency.
Citizens pay for government contracts. It is time they knew what they are paying for.
Africon. 2008. Unit costs of infrastructure projects in Sub-Saharan Africa. AICD background paper. Washington, DC: World Bank.
Bajari, Patrick, and Jungwon Yeo. 2009. Auction design and tacit collusion in FCC spectrum auctions. Information Economics and Policy 21 (2): 90–100.
Bandiera, Oriana, Andrea Prat, and Tommaso Valletti. 2009. Active and passive waste in government spending: Evidence from a policy experiment. American Economic Review 99 (4): 1278–1308.
CBI. 2014. Competitive accountable transparent: A value driven public services sector. Report. http://www.cbi.org.uk/media/2607468/cat_report_final.pdf.
Chever, Lisa, and John Moore. 2013. Discretion and efficiency in public procurement: Evidence from France. EPPP Discussion Paper No. 2012-09. IAE-Sorbonne, Paris.
De Silva, Dakshina G., Timothy Dunne, Anuruddha Kankanamge, and Georgia Kosmopoulou. 2008. The impact of public information on bidding in highway procurement auctions. European Economic Review 52 (1): 150–81.
De Silva, Dakshina G., Georgia Kosmopoulou, and Carlos Lamarche. 2009. The effect of information on the bidding and survival of entrants in procurement auctions. Journal of Public Economics 93 (1): 56–72.
Green, Will. 2014. Less than transparent. Supply Management. http://www.supplymanagement.com/news/2014/less-than-transparent.
Hendricks, Ken, R. Preston McAfee, and Michael A. Williams. forthcoming. Auctions and bid rigging. In Oxford Handbook on International Antitrust Economics, edited by Roger Blair and Daniel Sokol, Oxford: Oxford University Press.
International Bank for Reconstruction and Development/World Bank. 2011. Curbing fraud, corruption, and collusion in the roads sector. http://siteresources.worldbank.org/INTDOII/Resources/Roads_Paper_Final.pdf
Kenny, C. 2007. Construction, corruption and developing countries. World Bank Policy Research Working Paper 4271. Washington, DC: World Bank.
Kenny, Charles. 2010. Publishing construction contracts and outcome details. World Bank Policy Research Working Paper 5247. Washington, DC: World Bank.
Majeed, Rushda. 2013a. Promoting Accountability, Monitoring Services: Textbook Procurement and Delivery, The Philippines, 2002-2005. Innovations for Successful Societies case note. http://successfulsocieties.princeton.edu/publications/promoting-accountability-monitoring-services-textbook-procurement-and-delivery
Majeed, Rushda. 2013b. Power at the grass roots: Monitoring public works in Abra, Philippines, 1986–1990. Innovations for Successful Societies case note. http://successfulsocieties.princeton.edu/publications/power-grass-roots-monitoring-public-works-abra-philippines-1986-1990
Orlov, Eugene. 2011. How does the Internet influence price dispersion? Evidence from the airline industry. Journal of Industrial Economics 59 (1): 21–37.
Pan, Xing, Brian T. Ratchford, and Venkatesh Shankar. 2004. Price dispersion on the Internet: A review and directions for future research. Journal of Interactive Marketing 18 (4): 116–35.
Parliament of Victoria (2000) Contracting Out of Government Services http://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Finance_and_Public_Administration/Completed_inquiries/1996-99/outsourcing/report/c07
Pesendorfer, Martin. 2000. A study of collusion in first-price auctions. Review of Economic Studies 67 (3): 381–411.
Savedoff, William D. 2008. The impact of information and accountability on hospital procurement corruption in Argentina and Bolivia. U4 Brief 2008(7). http://www.u4.no/publications/the-impact-of-information-and-accountability-on-hospital-procurement-corruption-in-argentina-and-bolivia/
Sharing the spoils. 2014. The Economist, September 9.
Sistema Electronico para la Contratacion Publica. 2012. Sistema Electronico para la Contratacion Publica http://www.alcaldiabogota.gov.co/sisjur/normas/Norma1.jsp?i=28776.
Strassmann, W. Paul, and Jill Wells. 1988. Global construction industry: Strategies for entry, growth, and survival. World Industry Studies. London: Unwin Hyman.
Transparency International. 2013. Simplified procurement: Corruption risks in non-competitive government contracts. http://transparency.ge/en/node/3667
United Nations Commission on International Trade Law. 2012. Model law on public procurement and guide to enactment. http://www.uncitral.org/uncitral/en/uncitral_texts/procurement_infrastructure/2012Guide.html
United States House of Representatives, Committee on Government Reform—Minority Staff, Special Investigations Division. 2006. Dollars, not sense: Government contracting under the Bush administration. http://oversight-archive.waxman.house.gov/documents/20061211100757-98364.pdf
Rights & Permissions
You may use and disseminate CGD’s publications under these conditions.