FOSS Government Policy/Strategies
There are numerous strategies for achieving any conceivable policy goal. Each strategy has its own advantages and drawbacks. This section discusses considerations affecting strategy formulation, and some of the more commonly used strategies with their advantages and drawbacks.
No strategy should be taken and implemented without careful consideration of the local environment. Issues such as language, economic development, legal environment and cultural attitudes can make a particular strategy impossible.
To date, FOSS promotion strategies via government procurement throughout the world fall into four broad categories. They are:
- Mandating FOSS
- Preferring FOSS
- Mandating Open Standards
- Best Value
- 1 Mandating FOSS
- 2 Preferring FOSS
- 3 Mandating Open Standards
- 4 Best Value
- 5 Level of Development
- 6 Policy and Legal Environment
- 7 Capacity Development
- 8 Research
- 9 FOSS as an Industry
- 10 Footnotes
This is the most radical approach as it mandates the usage of FOSS systems throughout the government sector. In some countries, this means replacing the entire existing proprietary infrastructure, which involves large implementation and training costs. Although a number of proposals (legislative or otherwise) to this effect have been submitted, to date few have passed. The high costs and risks involved are the main deterrents in this approach.
Other countries have chosen the less painful route of mandating FOSS for all new procurement. More conservative approaches such as this are somewhat more common. Countries that mandate the change of only a proportion of the infrastructure over to FOSS include Brazil (80 percent of all systems) , South Korea (20-30 percent of all systems)  and Thailand. .
This approach greatly promotes FOSS usage and capacity in the local economy. However, the criticism is that this is done at the expense of the proprietary software industry. Certain economically advanced countries have also criticized such policies as being protectionist and against the spirit of free trade. China has a policy of blocking of foreign software usage in government offices . This does not mandate FOSS per se but it has a strong stimulating effect. The policy considers locally packaged FOSS systems as local software, even if the international FOSS community produces the majority of its components.
Recognizing the difficulty of switching the entire government infrastructure over to FOSS, many governments have moderated their approach by preferring FOSS solutions for the new procurements. When all traditional commercial measures are equal (functionality, TCO, risks, stability, etc.) then the FOSS solutions are selected in recognition of the social benefits, which can be hard to quantify.
This approach has the benefit of being easier and less risky to implement. It is also more flexible, allowing procurements to be decided on a case by case basis, taking into account factors such as the possible lack of a local developer pool. However, the weaker mandate may not be enough to counter the advantage that proprietary software enjoy when there is an established proprietary system.
Preferring FOSS: South Africa
Sections 10.1 and 10.2 of South Africa’s OSS strategy states:
Government will implement OSS where analysis shows it to be the appropriate option. The primary criteria for selecting software solutions will remain the improvement of efficiency, effectiveness and economy of service delivery by Government to its citizens.
OSS offers significant indirect advantages. Where the direct advantages and disadvantages of OSS and PS (Proprietary Software) are equally strong, and where circumstances in the specific situation do not render it inappropriate, opting for OSS will be preferable.
The full strategy can be obtained from:
Mandating Open Standards
Mandating open standards often has a complementary effect on FOSS systems. One of the most effective ways in which software vendors lock in their users is the use of proprietary standards. FOSS systems are at a disadvantage in a mostly proprietary software environment due to the lack of interoperability. The mandating of open standards would level the playing field and introduce increased competition, not just between proprietary software and FOSS but also between different proprietary software solutions. However, this often requires modifying procedures and legacy documents that are still stored using proprietary standards.
The two areas often targeted by open standards advocates are documents and web standards. Text documents and spreadsheets are typically stored in proprietary formats and may not be retrievable without the proper proprietary software, thus impeding the free exchange of information. Proprietary, closed web standards are ironic, since the World Wide Web is primarily based upon open standards. However, the dominance of a single web browser and its complementary web development tools from the same vendor have resulted in many Web sites being created using non-standards compliant HTML tags that are only accessible using Internet Explorer even though it would take minimal effort to make these sites cross-platform.
Mandating Open Standards: Commonwealth of Massachusetts, USA
- Effective and efficient government service delivery requires system integration and data sharing.
- Technology investments must be made based on total cost of ownership and best value to the Commonwealth. Component-based software development based on open standards allows for a more cost-effective build once, use many times approach.
- Open systems and specifications are often less costly to acquire, develop and maintain and do not result in vendor lock-in.
- All prospective IT investments will comply with open standards referenced in the current version of the Enterprise Technology Reference Model.
- Existing IT systems will be reviewed for open standards compatibility and will be enhanced to achieve open standards compatibility where appropriate. Open standards solutions will be selected when existing systems are to be retired or need major enhancements.
The Commonwealth of Massachusetts’ Enterprise Open Standards Policy can be viewed at:
In some cases, mandating open standards would initially preclude certain proprietary software vendors from participating until such time that they add proper support for open standards in their products.
Emphasis on open standards is strongest in countries with mature ICT industries and infrastructure. The European Union, the United Kingdom, certain states within the United States and New Zealand  are among the governments supporting open standards.
This approach focuses mostly on the economic value of FOSS, de-emphasizing the national and social benefits of wide scale FOSS adoption. This approach is the least controversial and is the standard policy in most countries.
However, due to the relative newness of FOSS and lack of general awareness, there have been calls for legislation or policy that explicitly places FOSS on the same level playing field as established and reputable proprietary software. Multiple legislative initiatives were started at the state level in the United States but, to date, none have passed. Any policy or legislation that explicitly requires that FOSS be considered on an equal footing with proprietary software is strongly opposed by proprietary software companies and intensely lobbied against.
Level of Development
The first and most important factor to note is that the level of economic development and ICT infrastructure greatly affects the strategies employed. Depending on whether a country already has an existing infrastructure, appropriate strategies change significantly.
Relatively advanced nations typically have existing ICT infrastructure and a trained technical pool. The infrastructure and skill set are typically on proprietary systems and this poses significant problems in several ways:
- Institutional resistance
- Since decision makers are familiar with their particular version of proprietary software and their skill sets are all rooted in this system, any FOSS policy is likely to encounter strong resistance. Any institution is resistant to change and ICT departments are no different. In new procurements, decision makers are likely to be more comfortable staying with the same technology as their legacy systems and biased against new technologies, including FOSS systems.
- Migration costs
- Since legacy systems are already in place, implementing FOSS systems will eventually involve migrating existing systems over. Migration can sometimes be even more expensive than implementing a system from scratch. This is due to retraining costs and difficulties involved in migrating data from proprietary data formats or interfacing with legacy systems that were not built with interoperability with other systems in mind. Even in a minimally computerized environment, moving simple word processing documents from a proprietary data format to an open standard can be a significant amount of work .
- Even if a total migration to FOSS systems were contemplated, there will be a period of time when both FOSS and proprietary systems must coexist peacefully. Unfortunately, a common strategy with proprietary software is to make their systems not fully compatible with other systems, be they competitor proprietary software or FOSS systems. In some cases, this incompatibility is relatively minor (for example, text documents that are strangely formatted). But in other cases, data interchange may be quite complicated.
For these reasons, implementing a FOSS infrastructure or at least creating a FOSS-friendly environment is much simpler when the existing infrastructure is limited.
With relatively developed countries, FOSS policy issues tend to focus on migration strategies, open standards, new procurement and coexistence issues. Relatively undeveloped countries focus on capacity- building and legal issues to ensure that the infrastructure they build, from the very beginning, is FOSS-friendly.
Policy and Legal Environment
Most countries have an existing body of policies and laws. As the proprietary model of software development has been the predominant model until recently, existing policies and laws are normally favourable towards proprietary software. In some cases, this may mean that some policies or laws create a hostile environment for FOSS.
Many countries are finding that they have to re-examine and often tweak their existing policies and laws to align them with their FOSS goals. Often, the changes are not just within the ICT sector but within other sectors as well, such as copyrights, patents, industry promotion and education. For some of the more common issues, please refer to the “Cross-Sectoral Concerns” section later in this primer.
Countries that have directly addressed this problem, frequently form a working group or committee, to identify the different problems and work with the appropriate government body to address them. This often involves a process of awareness raising and education of policy makers in other sectors of the government. Obtaining consensus can often be a lengthy process and it helps to have support for FOSS from the highest levels of government.
Many developing countries have noted that a serious shortage of FOSS capacity in their economies holds back the implementation of major FOSS projects. Any major FOSS policy in developing countries will involve building a pool of local FOSS experts and companies to support the projects and users of FOSS in the country
There are several focus areas for strategies on FOSS capacity development:
The traditional educational structure, starting from primary schools up through to the university level, can often be an excellent training ground for FOSS. There are a wide number of strategies in this sector, too many to be listed exhaustively. Some of the more common strategies include:
- Computer labs with FOSS installed
- Projects such as these have been carried out in various countries, including Brazil, Mongolia, Spain, and Thailand. In some cases, the initial purpose was not to build FOSS capacity but to build ICT capacity in general. FOSS systems were utilized to reduce expenses but this also had the added advantage of introducing an entire generation of students to FOSS relatively painlessly. Mongolia, among others, has reported that students had no difficulty learning and using FOSS systems. Implementing an initiative such as this requires a minimum level of FOSS capacity within the country to support the school computers. Otherwise, some schools may find their computers not being used, as was experienced by the Goa Schools Project. 
- Ensuring that curriculum is vendor neutral
- Many ICT literacy and computer science programs in schools today, even at the university level, are written with a specific proprietary software suite in mind. By ensuring that the teaching of ICT concepts is decoupled from vendor-specific skills, a more level competitive field can be achieved. In most cases, ICT skills can be taught on many platforms. For instance, basic ICT literacy skills such as email, web browsing and word processing can be taught on multiple proprietary and FOSS platforms. It may even be beneficial for students to experience two different implementations of a certain concept (one proprietary, one FOSS) to ensure that the students learn flexible skills that can be easily transferred from one system to another. This does place an extra burden on both educators and students.
- Some universities justify their teaching of skills tied to specific software packages by noting that their graduates are likely to be more productive immediately upon graduation. However, educators and policymakers should keep in mind that most software systems have extremely short lifespans. A software package taught today is often outdated within five years and a graduate who has not learned broad, easily transferred concepts will become obsolete.
- Scholarships/innovation awards
- These are relatively small financial rewards given to students who contribute to the development of FOSS. Meant more as encouragement and recognition of their achievements rather than financial support, these awards raise awareness and interest in FOSS. This will in turn lead to more skilled FOSS practitioners within the formal education system. Such awards have been implemented in places such as India, but it is still too early to assess the long-term outcomes of such a programme.
Certification and Retraining Programmes
Most policies, especially those with short-term goals, normally have a certification and retraining component as well. It typically takes less time to retrain already skilled technical personnel than it takes to train new personnel from scratch through the formal education system.
Retraining programmes can either be part of continuing education classes taught in the evenings at colleges and universities or in specially mandated training centres. Some initiatives in the Asia-Pacific region have had government-subsidized FOSS training as part of the retraining benefits given to unemployed or downsized workers.
A certification programme is also recommended to complement the retraining effort. Certification ensures that all technical personnel possess a consistent and sufficient level of skill before they are allowed to implement FOSS projects. These minimum standards are necessary to ensure that initial FOSS projects do not fail due to a lack of sufficiently qualified personnel, thereby creating a bad impression for future projects.
An alternative to creating a certification programme from scratch (which is quite a significant undertaking) would be to utilize one of the globally recognized certification programmes. Two of the notable certification programmes for GNU/Linux (which is only a subset of the entire FOSS movement) are the Linux Professional Institute's Certification (LPIC) and Red Hat’s Red Hat Certified Engineer (RHCE) programme or the Comptia Linux+
- The LPI is a non-profit organization that administers a distribution-neutral examination. The LPI tests the most commonly used Linux skills via a multi-choice exam plus some fillin the blank examinations. The neutral rather than specific vendor approach has its merit. Two exams are required for Level One at a cost of $300(US)
Comptia Linux + is also operated by a non profit organization and it is a single exam. They updated the exam in recent years
BSD exams are also available priced at only $75 for the one exam which has employed a psychometrican
Pakistan is training up to 10,000 end users and 1,000 system administrators on FOSS applications and systems as part of its e-government training project. The training will be focused on employees of the Federal and Provincial Governments in the major cities.
Targeted at increasing the standard and quality of public services through the effective use of ICT in the public sector, this training programme is unusual in that it covers both proprietary (Microsoft Windows and Office) and FOSS (GNU/Linux, OpenOffice.org) applications. Trainees are then able to compare and choose between the different systems.
More information can be found at:
- This examination focuses primarily on Red Hat’s distribution of Linux but a good proportion of the skills can be easily transferred between different FOSS distributions. The RHCE also includes a hands-on practical examination involving the actual set up of a GNU/Linux system. The LPI does not offer a practical exam component at present. However, the RHCE is specific to a single Linux distribution and significantly more expensive.
Both organizations are willing to cooperate in localizing their examinations, so language considerations are not a major issue.
FOSS Competency Centres
A common strategy of many FOSS policies is the creation of a FOSS competency/research/compatibility centre that performs a variety of functions:
- FOSS awareness raising and promotion.
- Implementing pilot projects in e-government.
- Documenting current best practices of FOSS usage.
- Providing technical support for government agencies.
- Providing training of both end-users and technicians.
- Aggregating news and resources on FOSS.
These centres typically serve as the core of FOSS competency supporting the development of competencies elsewhere in the country, especially in the public sector. In larger countries, multiple centres may be established, one per major metropolitan area.
|FOSS Competency Centres: Malaysia|
|Malaysia's Master Plan for implementing FOSS in the public sector includes the creation of an Open Source Competency Center (OSCC). The target activities are listed below. Further details available from: http://opensource.mampu.gov.my/|
|Policies and Guidelines||
|Promotion and Awareness||
|Training and Certification||
|Research and Development||
Some policies include a requirement that software produced by publicly funded research should be released under a FOSS license. The rationale is that the fruits of public funds should be freely enjoyed by the society as a whole. Releasing it under a FOSS license would ensure this, as opposed to a trend in some countries to have publicly funded research patented and licensed to commercial organizations to commercialize the results of the research.
The United States has a situation that is similar in principle: most of the software created by US departments is released into the public domain, made available to all to utilize and take advantage of. As this situation has been in place for many years, some of the results are quite interesting. The VISTA  hospital information system released by the Veterans Administration is a widely used FOSS system that runs hundreds of hospitals around the world, saving millions of dollars in software costs in each hospital. The bioinformatics field throughout the world relies very much on software created by the National Center for Biotechnology Information (NCBI). 
In both cases, the primary beneficiaries are the citizens of the country itself, as adoption is normally quickest there. But the benefits have spread globally.
Some critics of such initiatives say that releasing research under a FOSS license is unfair to the private sector, as this would prevent the commercialization of this research. However, this criticism applies to only one or two of the 52+ FOSS licenses. There are a variety of licenses that are especially business-friendly and some were even created by corporations such as Sun, IBM or Apple. Deciding the appropriate license for research is best left to the local legal experts.
FOSS as an Industry
As a challenge to the traditional proprietary software industry, FOSS can be a large contributor to the local economy. Large multinational corporations such as IBM and Red Hat, as well as countless small and medium-sized companies such as MySQL AB, Digital Creations and Trolltech, have generated strong profits by focusing on FOSS.
While a large traditional proprietary software firm like Microsoft is unlikely with the FOSS model, a vibrant ICT industry based on FOSS is quite possible and several countries have already begun policy initiatives to encourage a FOSS industry.
In many developing countries there is insufficient local capacity to properly support widespread FOSS usage (or ICT usage of any kind) throughout the economy. Thus, FOSS policies can and should be linked with the creation and development of Small and Medium-sized Enterprises (SMEs) to supply support services. In combination with a FOSS government procurement strategy, this can create a vibrant ecosystem with strong supply and demand.
There are several strategies that governments may employ to encourage the growth of FOSS among SMEs, as discussed below.
SMEs normally do not have access to the wealth of information resources that larger corporations have and they may be unaware of the full benefits of FOSS. Awareness-raising activities such as conferences, workshops and training sessions will help SMEs fully utilize this resource and incorporate it into their organizational strategies.
Although government procurement has been covered as a general method of promoting FOSS within a country, it can serve as a vehicle to promote SMEs as well. Most government procurement contracts are too large for SMEs to compete for and this normally places them at a competitive disadvantage. However, some governments (notably the United States and to a lesser extent the United Kingdom) have requirements that a certain percentage of a government contract must be carried out by SMEs or be subcontracted out to SMEs by the ultimate winner of the contract.
These initiatives have created strong partnerships between large corporations and SMEs, and have helped quite a few SMEs grow to a size where they are able to compete with the larger corporations in their respective countries.
Besides directly creating demand via overt purchases, governments can increase the attractiveness of FOSS to SMEs by offering incentives such as tax breaks. Countries like Singapore have offered tax reductions to companies that use the GNU/Linux operating system.  The guaranteed cost saving makes FOSS systems more attractive to SMEs.
In rare cases, governments can choose to support local FOSS companies via loans, loan guarantees, seed funding, venture capital funds and other financing methods. Assuming there is sufficient demand in the local economy for FOSS products and services, government support can reduce the financial roadblocks to starting and running a successful FOSS business, particularly among SMEs.
Many countries have credit programmes for small businesses. An example is the United States’ Small Business Administration (SBA). The SBA has put over US$30 billion in the hands of small business owners since 1958 to finance the growth of the United States’ vibrant SME industry.  A modified version of such plans can be adopted to assist FOSS SMEs.
- “The Brazilian Public Sector to Choose Free Software”, PCLinuxOnline, 2 June 2003; available from http://www.pclinuxonline.com/modules.php?name=News&file=articleamp;sid=6879
- Myung, S. E., “Korea Jettisons Windows for Linux”, silicon.com, 1 October 2003; available from http://www.silicon.com/software/os/0,39024651,10006225,00.htm
- “ICT Ministry Sets Linux Targets for government Sector”, Bangkok Post, 23 June 2003.
- “China Blocks Foreign Software Use in Govt”, CNETAsia, 18 August 2003; available from http://asia.cnet.com/newstech/applications/0,39001094,39146335,00.htm
- Brislen, P., “Microsoft Defends XML Patent, Offers Royalty-Free License”, Computerworld.com New Zealand, 5 February 2004; available from http://computerworld.co.nz/news.nsf/NL/AB73832F819E0AF5CC256E30000C7BFE
- Expressed during the Free and Open Source Software Asia-Pacific conference held between 9-11 February in Kuala Lumpur, Malaysia. Summary findings and recommendations are available at http://www.iosn.net/downloads/fossap-report-final.pdf. A second conference in Africa expressed similar need. The Idlelo conference was held on 11-15 January in Cape Town, South Africa. The final report is available at http://www.fossfa.net/tiki-download_file.php?fileId=3
- Martyris, D., ”Community -Government Partnerships and Open Source Technology for Low Cost IT Access in India – A Case Study”, July 2003; available from http://www.developmentgateway.com/node/133831/sdm/blob?pid=5474
- More information available on the Internet at http://www.worldvista.org/vista/index.html
- The basic algorithms that are used in biotechnology research are based on NCBI’s tools. Although some companies have re-implemented the algorithms, most use the existing algorithms and add value on top of it. Further information can be found at http://www.ncbi.nih.gov
- UNCTAD Secretariat, “e-Commerce and Development Report 2003”, 2003, United Nations Conference on Trade and Development; available from http://www.unctad.org/en/docs/ecdr2003_en.pdf
- “Overview & History of the SBA”; available from http://www.sba.gov/aboutsba/history.html