New Protectionism: Mandates for Open Source Software

America has always been a trading nation. From the beginning, access to foreign markets was critical for the development of commerce. Today, international trade plays a vital role in the American economy, providing consumers access to low cost goods from around the world, while providing American producers a global market for the goods they produce. In June, Americans imported $124 billion in goods and services while exporting more than $84 billion goods and services. Roughly 12 million Americans are employed in export-related jobs.

For consumers across the globe, the benefits of open and competitive markets are clear: lower prices and a wider choice of goods and services. Producers, on the other hand, are often threatened by low-cost competitors, and seek political protection against outside competition. Tariffs and quotas are the more traditional methods of keeping out foreign competitors. However, non-tariff barriers are a more subtle tool that permits governments to talk of free trade while serving local interests. Restrictions based on technical standards or government procurement policies protect domestic producers just as effectively as a tariff.

In the technology field, there have been efforts around the world to adopt “open source” mandates for all government software purchases. Open source refers to software whose underlying programming code is not proprietary; it is available to all users to modify and adapt to their own needs. Linux, an operating system that competes with Microsoft Windows, is perhaps the most well-known of the open source products available in the market today. By contrast, proprietary software relies on copy-protected code that is licensed for end users. Microsoft, Oracle, and Adobe are common proprietary software producers.

Starting in Brazil, the open source movement has expanded to more than 40 countries around the globe. From Malaysia to France, nations are aggressively pursuing mandates to promote or favor open source software. Such mandates often are defended using the same tired protectionist arguments that have been used for generations by governments attempting to keep foreign products out. A common argument is the “infant industry” theory. Namely, a country has to protect local industry so that it may grow to the point where it can be competitive in a global market. More often than not, however, such claims are not made by fledgling industries; they are made by those industries that have the political influence to thwart competition. Assertions that every nation needs its own software industry are just as dubious as earlier claims that every country needed an automobile industry or a steel industry. In fact, gains from trade arise from nations specializing in producing what they do best.

This is not to say that all software need be proprietary. Decisions about software purchase should be left to the market, allowing all producers—open source and proprietary—to compete for customers. Products should be evaluated based on the value they provide to the end users. This must be compared to costs of using the software, which incorporates not only the upfront purchase of the software, but also other costs, such as the cost of customizing the software to fit the needs of the end user, the costs of debugging the software, the costs of maintaining the software, and so forth. In a study for the AEI-Brookings Joint Center for Regulatory Studies, Professor Alan MacCoramack of the Harvard Business School finds that acquisition costs account for less than 10 percent of the total costs, which means that simply comparing the upfront costs of proprietary software to the “free” price of open source software is a poor measure for evaluating software options. Consequently, mandating the use of one over the other does little to ensure lower costs or a better product. It does, however, establish a non-tariff barrier that would limit the ability of American software companies to compete in the global marketplace.

Just as concerning, open source mandates are being pursued in several states as well. Oregon and Texas, for instance, are considering open source mandates for government purchases. Locking governments into one system limits competition and may generate higher costs for the taxpayer. This is especially true of “open source” mandates, because there may be more incompatibility problems among open source programs than proprietary programs. The software industry in the United States is highly developed, so there should be little need for mandates when making decisions about software. The market is competitive, and end users—governments as well as private firms—have full access to a wide array of software. Mandates can only limit choices within that market.

Ultimately, a mandate for open source software should be viewed as any other mandate—it is an attempt to subvert a market outcome through government intervention. The fact that that governments rather than private firms (other than open source providers) are pushing the mandate only raises more questions, with an implicit suggestion that without government assistance, open source software cannot compete. Does this mean future subsidies to foster the open source industry? Open source software has made inroads into the server market, and can be evaluated as a legitimate alternative to proprietary software. Consumers constantly make these decisions, and the more products available, the better the chances they can find one that meets their needs. All that is needed is a market, not a mandate.