ICT-related procurements in the education sector, especially large scale ones, are not easy. A recent World Bank Internal Evaluation Group report  noted that "ICT procurement has been highlighted as a major implementation constraint in several country and regional portfolio reviews and is a critical dimension of design." Rapid changes in technology mean that many ministries of education have a hard time keeping up with what's current in the market, let alone what might be coming next.
Even in places where anti-corruption measures are well considered and implemented, government auditors and external watchdog groups may be challenged to identify dodgy practices in some ICT-related areas. (Have you ever read the fine print on large scale bandwidth contracts for schools? Such things are often not for the feint of heart.) It is not unknown to hear whisperings about vendors -- or consultants close to them -- providing 'assistance' of various sorts in writing a request for proposals (or certain technical specs that appear in such RFPS), and of course vendors often hope that their showcase pilot projects may inspire ministries of education to think in certain ways about what is possible, and even desirable. For many ministries of education, the line between 'influence' and 'undue influence' in such cases can be very clear in some circumstances, but rather hazy in others.
As part of a very interesting Q&A period after a presentation at the World Bank a few years ago, mention was made about some of the challenges faced in a state in southern India which was exploring whether so-called thin client solutions might be worth considering in its schools. Essentially, the issue was this: Traditional practice when procuring computers for schools had focused on ensuring that each computer met a defined set of minimum technical specifications. In an alternative, 'thin client' set-up, it was possible to use workstations that had less robust specifications, provided they were connected to a powerful server whose processing power substitutes for that of the client computer. To oversimplify:
[-] 'traditional' approach: lots of pretty powerful computers
[-] 'alternative' approach: lots of relatively underpowered computers, connected to one very powerful computer
The point here is not to imply that one type of arrangement is on its face better or worse. Rather, it is to highlight that, if you write an RFP in a certain way -- in our example here, requiring that *every* computer meet a certain relatively high technical specification (processing speed, hard drive size, etc.) -- you may exclude proposals that feature non-traditional or 'alternative' or new approaches.
One way around this is to put more emphasis on functional specifications, rather than technical specifications, in certain components of your RFP. Not sure what this means in practice? When discussing such issues with ministries of education, I often point to an RFP at the heart of a procurement process in the U.S. state of Maine as a way to highlight an approach to procurement that is, at least in terms of most of the places where the World Bank works helping to advise education leaders, rather rare. While I am certainly no procurement expert -- thankfully we have plenty of very good ones at the World Bank to whom I can refer people -- I offer the comments below based on many discussions with ministries of education about their challenges in this regard, in case doing so might be of any interest.
The Maine Learning Technology Initiative (MLTI) has been seen as a pioneer in exploring so-called "1-to-1 computing", where each child has his/her own computing device. (For more information,, please see a previous blog post, The Maine thing about 1-to-1 computing , or visit the MLTI  or Maine International Center for Digital Learning  sites.) As a result, Maine, despite being a small U.S. state, has had an outsized influence on thinking on educational technologies matters in many places.
I sometimes share Maine's 2006 RFP  [pdf, full text] as a way to provoke discussion around approaches to planning for, and purchasing, not just individual pieces of hardware, software and related training and support, but rather a whole integrated package of such things in ways that may help a ministry of education meet its objectives.
Maine will be releasing its next big RFP in December, and has been soliciting public input and feedback this year via a number of channels, including a dedicated web site . The good folks in Maine explain their approach to RFPs in this way:
"The RFP process is designed to describe a challenge and the desired outcomes. It is intended to provide specifications and parameters for the solution that may be proposed by a bidder. Finally, it is intended to be allow bidders as much flexibility as possible to propose innovative solutions to help solve the challenge and support the outcomes as defined by the RFP. This means that an RFP should avoid being prescriptive (i.e. defining what equipment or software tools should be included)."
Most RFP processes related to educational technologies that I have looked at are all about being prescriptive, about defining -- often to a very exacting degree -- what equipment or software tools should be included. In Maine, they are (roughly speaking) happy to leave it for the market to propose specific technical specifications, which are then assessed during the evaluation of the proposals that are submitted. Such an approach is potentially risky. It pre-supposes that there is a vibrant marketplace of firms who can propose workable, innovative solutions to meet the specific objectives that the government has identified. It also assumes that there is a highly competent set of people who can evaluate -- using a set of clear, rationale criteria -- proposals that could potentially feature radically different approaches to meeting the government's stated objectives. On the other hand, might it be more risky for the government itself to assume up front that it can accurately predict what the best specific technology 'solution' might be, especially in an area where technologies are changing so quickly?
Some very quick examples of language from the 2006 Maine RFP:
184.108.40.206 Device Portability 
What the RFP says: "The device will be able to be carried conveniently and easily by students and teachers – either via a provided carrying case or some built-in carrying ability. The portable computing device shall be lightweight. While the Department will not mandate a specific maximum weight, as a guideline the Department would prefer to see a device and all its components that weighs six pounds or less. In general, the lighter the better."
What the RFP does not say: The RFP does not identify the specific type of device, i.e. laptop, 'netbook' or tablet. Instead, it offers some guidance about what the attributes of such a device might be.
3.5.2 Device Reliability 
What the RFP says: "The solution will provide device reliability and a service level that ensures no student is without a functioning device for more than one (1) school day. This may mean that different support plans need to be in place for different schools."
What the RFP does not say: The RFP does not mandate a specific approach to solving technical problems with the devices. Instead it says that, whatever the problem is, a student can not be without a device he/she can use for more than 24 hours, and that the vendor is responsible for coming up with creative solutions to making sure this objective is met.
220.127.116.11 Services by Provider 
What the RFP says: "The bidder should describe the full potential for curriculum integration and system capabilities within the application of the proposed wireless network in the educational setting; describe how it would assist schools in identifying and achieving their desired level of curriculum integration and system capabilities; and describe the bidder’s experience in maximizing student achievement with wireless networks in educational settings."
What the RFP does not say: The RFP does not mandate a specific approach to the use of the vendor's equipment, content, etc. to help meet learning objectives. Instead, challenges the vendor to identify how its 'solution' will work, and how this will make a difference for student learning.
The examples are meant to illustrate how an RFP can articulate some specific functional requirements while at the same time challenging a vendor to demonstrate how it will help educational authorities and institutions meet their specific objectives. (For more examples -- including perhaps some better ones -- you are invited to have a look at the language of the RFP  itself.)
Even assuming (a) that an RFP is well crafted, informed by a clear vision for what a ministry of education wishes to accomplish; (b) that a sufficient number of competitive proposals are generated in the market; and (c) that government is able to evaluate such proposals according to a systematic, rational process; governments still need to negotiate and write a contract to guide the implementation of the project. In doing so, they need to ensure that they can effectively monitor the actions of the vendor, with clear penalties if certain objectives are not met, and provide for enough flexibility that, if things simply aren't working out or if technological innovations result in new opportunities for reaching project goals more effectively, both sides can come together to restructure things in ways that are legal, transparent and practical. Translating the grand ideas and visions and that animate proponents of doing lots of transformative things with technology in the education sector into concrete, measurable, workable contracts is no easy thing.
Many large scale educational technology initiatives are billed as 'public-private partnerships'. While I am not expert on such things, I do find that the 'partnership' component of many such initiatives is rather ill-defined. For me, PPPs are at their heart about the sharing of risks -- if the vendor is not sharing risks, what you have is really a basic contracting relationship, perhaps with a little corporate social responsibility thrown in to sweeten the sauce. There is nothing wrong with this approach, but it can at times be difficult to characterize such an arrangement as a real 'partnership'.
My former World Bank colleague Charles Kenny has been talking recently about "The Case for Routine Publication of Government Contracts ", arguing that, while public monies are used by governments to buy things, "the contracts that say who is going to provide what services, where, when, and how; those contracts aren’t public so citizens don’t know what they’re paying for.” Whether or not you buy Charles's argument, disclosing -- and archiving in a public, permanently accessible place -- RFPs related to government educational technology procurements would be an important first step toward greater openness and transparency. We are regularly asked here at the World Bank for examples of RFPs so that other countries can learn from them. Maine has done a good job of archiving  the RFPs it has used, as well as other key documents that have informed its history. You might wish to have a look; there's some interesting stuff there. Most other high profile educational technology initiatives around the world have not done quite as good a job of this. It would be interesting to see what might happen if some of the related contracts that results from these RFPs were made publicly available as well.
some related resources that might be of interest:
World Bank's procurement site provides some specific guidance related to Information & Communications Technology Procurement , including preparing standard bidding documents  and a useful short guidance note  [pdf] from the Europe and Central Asia department. You may also find this document on procurement arrangements applicable to Public-Private Partnerships (PPP) contracts financed under World Bank projects  helpful in highlighting a number of key issues.
NASPO's Comparative Review of State IT Procurement Practices  [pdf] in the United States may help places benchmark their own approaches to such things.
please note: The image used at the top of this blog post of a marriage proposal in Helgoland ("should she be more interested in ensuring that he meets her technical or functional specifications if this partnership is to work?") comes from a painting by the Canadian genre artist Henry Ritter via Wikimedia Commons . It is in the public domain. The second image of a group of people in Maine preparing some lobster ("at first glance things may look a bit strange when you see how they've been cooking things up in Maine, but you might find the approach there rather interesting ...") was taken by 'schooner guest' and originall uploaded to the Internet by 'CaptBren'; it also comes via Wikimedia Commons  and is in the public domain.