Public AdministrationWritten evidence submitted by Open Source Consortium (PROC 5)

A submission from the Open Source Consortium (OSC) a UK trade association for SME suppliers of services based on Free and Open Source Software (OSS) seeking to promote the advantages of open standards and OSS in an information society and knowledge economy.

Introduction and Summary

Our response is mainly directed to to aspects of procurement affecting SMEs and information and communication technology (ICT) with specific reference to open source software (OSS) and the supply of services based on OSS.

One of the reasons that the OSC was formed in 2006, was to explore the creation of consortia to bid for large contracts. At that time we were advised by relevant parties in the public sector1 that if we didn’t form consortia we wouldn’t win contracts.

The overall problem we identify is that of signalling in situations with asymmetric information.2 Too often we have found it is not a question of what is being said but what is being done. In general we recognise that it might not be easy for large bodies to transact with SMEs however limiting contracts to £100 million as the Minister for the Cabinet Office announced doesn’t create many new opportunities.

For this reason individual members of the Open Source Consortium (OSC) have tended to steer away from getting involved with large tender exercises, particularly in the public sector.

Observation and experience suggest that whatever the government says3 to suppliers, the past remains the best guide to the future.

What basis do we have for assuming that next time it will be different?

Advice given to departments4 makes it clear that what ever is being required of them

[legislation coming into force on 31 January 2013] places a requirement on commissioners to consider the economic, environmental and social benefits5 of their approaches to procurement before the process starts

it does not apply to services (eg, services based on OSS as OSS licences are most usually available at zero-cost)

The Act will apply to framework agreements. When procuring an above threshold framework agreement for public services procurers and commissioners must consider the [legislation] [except it] does not apply to services contracts awarded by calling off from a framework.

In any event, in advance of this legislation coming into force two large government departments have just signed enterprise agreements6 with one of the standard proprietary software vendors. These arrangements cite money savings which would seem to depend on how the savings are calculated.7

Detail

We give three examples of the signalling problem.

1. ADDSS

When the ADDSS framework was put out to tender,8 we noted at the time that this £4bn framework exercise was being conducted before Cabinet Office had concluded its work on defining an open standard. We did not consider this to be a positive sign.

In early July 2012 we received an email9 from a body that had been working with Cabinet Office on OSS and public procurement:

I expect you and OSC are light years ahead of me on this but for completeness I append the details as I have been asked to encourage as strong a response from the Open Source Community as possible. Tenders due by 8 August 2012.

This appeared to be a signal that there were opportunities. Given the time scales we had to move quickly or not at all.

We coordinated a decision and eleven members of the OSC agreed to take forward a bid.10

We raised funds from members to pay for the bidding process and engaged a suitable consultant to assist us with maximising our chances for success. We expended considerable time and effort. Misallocation of these two commodities is a major source of opportunity cost11 to SMEs.

In practice there were many reasons why this tender exercise was not suitable including but not limited to:

Heavy upfront process—SMEs do not have dedicated marketing, contract and legal departments, simply engaging with the tender required considerable time and effort. Pleas by all SMEs for a time extension were ignored in the later stages of the procurement by the government procurement service (GPS);

Sheer size and complexity of the tender—this appeared to be a number of different tenders combined into a super tender (Application Development, ERP, Information Management, Content Management and Security Consultancy) which probably suited the government requirement for efficiency and, as a result, benefited the large as opposed to SME community irrespective of it being seemingly broken into multiple lots;

Over emphasis on essay writing skills—Part B of the Award Stage became a huge essay writing competition (approximately 10,000 words), which benefit large marketing departments. None of these questions favoured such advantages that SMEs might be expected to bring (eg flexibility, responsiveness, innovation);

Inconsistent use of CPV codes—the lot descriptions showed some marked inconsistencies with the accompanying CPV codes, with 35 lots and up to 30 CPV codes per lot this was a huge pieces of analysis for SMEs to undertake;

The tender showed few signs of being OSS friendly for reasons that included:

The tender document format—an inbuilt assumption that the respondents were set up for Microsoft Office (.docx, .xlxs).12 The heavy formatting of the documents rendered them impossible to open in Libre Office. Requests that the documents be supplied in the older Microsoft formats.13 doc, .xls were ignored, causing us to expend time and effort before we could evaluate our ability to participate.

There was no mention of looking for OSS based solutions and where anything were mentioned it was proprietary—Lot 24 (Application Support and Maintenance) expressly mentioned Oracle, SAP, Siebel, JD Edwards, PeopleSoft and IBM

Software licence costs and flexibility—the criteria for the award of call-off contracts (Schedule 6 of the Framework Agreement) was devoid of any guidance on how to apply life-cycle costs (an area where OS can compete very effectively with proprietary standards).

We decided to withdraw from the tendering process as a damage limitation exercise which was more prescient that we realised as ADDSS was subsequently cancelled by Cabinet Office.14

2. Treatment of OSS by Cabinet Office

We were not directly involved in this example from late 2011. A newspaper article informed us that:

“[A Cabinet Office official] said that simply considering open source alternatives helps improve competition, and mentioned that he encouraged a department to pilot open source LibreOffice as an alternative to upgrading its Microsoft software. This led to Microsoft providing the new software for free.”

We made an FOI request15 to find out more as on the face of it the example breaches EU state aid rules. The FOI reply claimed that the newspaper article was:

[…] not an accurate account of what the Cabinet Office official said at the EHI Live 2011 event. […] The Department has contacted the [newspaper] about this inaccuracy.

We contacted the newspaper and received this reply from the section editor16:

[…] I’ve checked back on this and can tell you that we did make a change, but it was only to correct a misunderstanding over [the official]’s position.

We originally reported him as being from CESG, but we received an email from their press department telling us he is a Cabinet Office official and made the relevant change. Nobody has questioned what we reported him as saying at the event.[...]

3. An NHS Scotland pilot project for an OSS desktop

The outcome of this exercise was a decision to enter into a new enterprise agreement for Microsoft software.17

(we note also that both the Department for Education and HMRC18 have just signed new enterprise agreements for Microsoft software.)19

OSC members formed a consortium to bid for this. On investigation we discovered requirements:

to use hardware capable of supporting Microsoft Windows 7;20 and

to use Microsoft Internet Explorer21 for which every instance requires a Microsoft Windows OS licence.

Accordingly the cost and complexity of any solution in effect precluded sensible use of OSS.

January 2013