Free Can Mean Big Money: The Open Source Economy

I read something in one of the comments for an OSNews posting a couple weeks ago that sent me thinking. It wasn’t an original or profound thought. In fact, it’s a rather commonly-held opinion that happens to be quite misguided. It’s an opinion summed up by the “open source = communist” meme that gets thrown around in thousands of flamewars all over the internet. In this essay, I will explore why this idea is wrong and demonstrates a fundamental misunderstanding of economics.

The following was posted in an OSNews discussion in July 2004:

I guess free software foundations are going to employ people from now on. Its the same evil mega corporations that employ hundreds of thousands of people and make the world economy function. Make them “smaller, weaker, and easier to keep in their place” and raise the unemployment rate to double digits not to mention lowering the standard of living world wide I suggest voting NO for RMS Democracy.

In other words, the money that is made and the jobs that are provided by the licensed software industry are an essential pillar of the economy, and any challenge to the status quo would have catastrophic effects. This misconception is actually rather easy to debunk, but it’s related to a more serious notion that merits serious discussion: the idea that replacing the now-dominant intellectual property regime with one that favors, or even enforces, sharing rather than hoarding is a threat to the world’s economic well-being.

An examination of the facts, put in historical perspective, shows that the engines of global progress have always been fed by the sharing of knowledge. In fact, if knowledge about, say, new agricultural techniques, like irrigation, had been hoarded and protected from competition, it would have set back the rise of civilization by centuries. It was precisely because early pioneers shared their knowledge (willingly or not) that the march of progress led steadily on.

This sharing was, in earlier times, unavoidable to some extent. Early innovators would certainly have been eager to maintain profitable monopolies on their ideas if there had been a mechanism to allow it. But the invisible hand of the free market applies a constant, inescapable pressure on idea-hoarding. In fact, the right to compete by producing a similar or identical product to another is one of the cornerstones of capitalism. Copyright and other protections of intellectual property are actually anti-liberty, anti-capitalist notions, though all but the most radical libertarians would recognize that measured protections are essential to promoting economic progress.

As in all things, what’s good in moderation can be harmful in large doses (or if withheld altogether), and overzealous protection of intellectual property stifles innovation in the long run. In an ironic twist, there is a type of economic system in which an organization is granted a right to be the sole producer of a particular good, protected from competition. That’s the “planned economy” model embraced by Soviet Communism.

Open source fanatics are communists. Just as in public discourse, all topics of disagreement seem to eventually degrade into someone calling someone else a Nazi, disparagement of the free software movement seems to be inescapably drawn to a comparison with communism.

The quote I highlighted earlier doesn’t explicitly call open source proponents communists, as such statements often do, but it does ascribe a foolish, perhaps inadvertent, anti-capitalist, anti-progress bent to open source philosophy. Now, to be fair, some open source proponents are anti-capitalist. Many are socialists at heart, and some may even be bona-fide communists. But even though extremists on both sides of the issue might stress (for positive or negative purposes) that open source software supports an anti-capitalist agenda, carefully considered evidence just doesn’t support the claim, and the non-extremists that make up the majority should reject that characterization vigorously.

Let’s dwell for a moment on the “communistic” aspect of the free software ideology. Again, I will not deny that some proponents of free software do, in fact, share some ideological common ground with Communist thinkers. For the sake of clarity, let’s leave the failed experiment of Soviet “Communism” out of this for a moment and focus on the theoretical (and apparently impractical) ideas proposed by Marx and other early 20th century philosophers.

So the radical fringe of the free software movement, Richard Stallman being the most prominent, can somewhat fairly be compared with the Communists of the 1930s

According to my understanding, the essence of philosophical Communism is that modern history is defined by lopsided power relationships, with a large poor class oppressed by a rich ruling class, and these groups are in struggle. Communism claims that society is evolving, with some kind of historical inevitability, toward the common person having more freedom and power. In this view, capitalism was an incremental improvement over Feudalism, allowing some of the oppressed to rise up and become oppressors themselves, but it will give way to Socialism and eventually Socialism will give way to Communism in some sort of inevitable progression. Control of the working class by a moneyed elite will be supplanted by a benevolent caretaker state that will enforce equality and grant power and freedom to the common people, and eventually those people will be able to administer to themselves, the state will cease to be necessary, and everyone will receive according to their need. All work will be done for the good of the community through an enlightened volunteer effort. Thus the continual class struggle will end in a kind of worker’s utopia. It all sounds pretty unlikely to our modern, jaded sensibilities, but back in the 1930’s I guess it sounded like it was worth a shot, since the transition to Capitalism had brought about terrible suffering in much of the world.

What really happened, of course, when this philosophy was put into practice was that the state that was empowered to administer this glorious change found it difficult to enforce these ideals on the common people it was supposed to help. So it established an authoritarian machine to force the ideals on them “for their own good.” That machine almost instantaneously became corrupt because of humanity’s love of power, and the state that was supposed to wither away only became stronger and more authoritarian until it collapsed under its own weight.

The more radical elements of the open source software “movement” share an important element of this philosophy: utopianism and a belief in people’s willingness to volunteer their labors for the common good. And like the early Communists, these people may have initially been driven by a naive view of human nature. The truth is, well-educated software engineers with good earning potential aren’t going to dedicate hours upon hours of time for some idealistic pipe-dream. Just as the people in the “worker’s paradise” of Communism turned out to not be so interested in “volunteering” their toil in the factories and mines so that a bunch of freeloading intellectuals in the cities can receive food and housing “according to their needs” while they compose sonnets about the valiant struggles of the working class.

If the nascent free software movement had turned out to be all about a utopian vision of sharing and pretty flowers, it would have gone nowhere. Unlike communism, free software was not about a life-controlling government, and Richard Stallman was never appointed dictator, so nobody could ever be forced to take part. Why did people do it, then?

The software industry actually has experienced unending class struggle just as world politics has. it’s a struggle between the scientific/intellectual class that produces most of the world’s software and the financial/managerial class that runs most of the companies that fund, market, and sell much of the world’s software.

Software producers generally believe in an academic-inspired ideal that ideas are best cultivated in an open environment with peer-review and researchers “standing on the shoulders of giants” by learning about and improving upon others’ ideas. More knowledge and better technology are the goal, and this goal is achieved by advancing the state of the art. They have their own strict protections of intellectual property, mostly involving a policy of never taking credit for work that isn’t yours by meticulously attributing all of your sources.

The financial/managerial class has its own value system, based mostly on the necessity to monetize the company’s assets. Firms have a responsibility to maximize the return on their investors’ money, so every company asset must be leveraged to its utmost. This means that if you have developed a program that can be sold for $1,000,000 to four people in the world or $100 to three million people, it is your solemn duty to keep the price at $1,000,000, even if that means that 2,999,996 people who need that software will have to go without. And at that price, you must keep your company knowledge absolutely secret, advancing the state of the art be damned.

Often these conflicts do not bump up against each other too much. The software producers need an environment in which they can create software (they need to be paid, be provided with desks, computers, etc) and the managers need the producers to have a product to sell. It’s a symbiotic relationship. But engineers often bristle at management’s lack of interest in funding inventive new research and instead packing useless bells and whistles into the existing products because sales and marketing think it will help make more money. And managers often decry programmers’ love of technology for technology’s sake and seeming lack of interest in the financial well-being of the firm.

Members of the scientific/intellectual class looking in from the outside have historically been disappointed that so much of the fruits of the labors of the engineers working within industry have been locked away from them. Either they can’t even afford to use the software because its licencing fees are so high, or if they can use it, they can only participate as an outsider because the intellectual property is guarded so carefully. If you’ve got some great ideas on how to make Microsoft Word better, you don’t have much recourse other than penning a letter to the product manager at Microsoft.

Unless, of course, you’d like to take a crack at writing your own word processor. And that is exactly what many of those disenfranchised members of the scientific/intellectual class did.

As an example, AT&T, due in part to its status as a regulated monopoly, was quite generous in letting people use Unix, and it inspired a lot of smart people to advance the state of the art. But the Unix OS and the proprietary utilities it needed to be useful were owned by corporations, and while it’s one thing to let academics use it, it would have been reckless to let them make new, commercially viable versions, especially if those versions were clearly better than the original. Shut down on that front, some intrepid programmers decided to re-write these utilities, and eventually the whole OS, from scratch. There were even competing versions of the re-writes, and camps formed around which was better. Over the years, others improved upon those re-writes, sometimes forking off into new projects. So in a way, the principles of the free market were applied to the production of a single piece of software, not just to sales and distribution of similar products.

How was it possible for these programmers to build upon each other’s work, to fork off competing versions, and pursue diverging philosophies of development? It was done using a tool that had been invented by the financial/managerial class: intellectual property licensing. But the licenses that these people used were different. Their aim wasn’t to monetize the software, but to strike a balance between enforcing the original author’s rights while encouraging an academic, collaborative sharing of knowledge. That original author might want only to be recognized for his or her work, or might want to require that derivative works must also be released under an equally open license.

Now for the academics and intellectuals involved, all of this effort was a slam dunk. They got access to more and better software than ever before. If they needed a system to do something special, they need not necessarily reinvent the wheel. There may be some software available that gets them 75% there, and they must only make the necessary improvements, saving time and precious resources.

And this is the way it stayed for a while. It was mostly academics enjoying the fruits of their collaborative labors. Industry wasn’t too worried. They had mostly provided software to these folks for free or for heavy discounts anyway, and there were plenty of good ideas and useful little tidbits of software coming out of this movement.

So there was a certain amount of idealism involved, though it was a sort of scientific idealism and an aversion to the kind of wasted parallel effort that the competitive commercial process naturally engenders. But it was far from being anti-capitalistic. In fact, it allowed ideas to compete in an open marketplace, not just in the enclosed sandboxes of corporate-sponsored R&D facilities.

But let’s not denigrate the sandboxes. The ideas that had been coming out of places like Bell Labs and Xerox PARC were pretty amazing. In fact, if the open source movement has taught us one thing, it’s that there’s really no substitute for the kind of sponsorship of R&D that’s done by corporations, academic institutions, and governments. What we learned, though, was that a lot of the dynamism that was coming out of these fertile nesting grounds was never making it into profitable commercial products, and ended up lying idle. That’s generally a source of great frustration for the software producing class. In fact, even most software that ends up working great and being useful never ends up being a profitable product, because most software is actually not written to be sold but is for internal use in a particular company or institution.

This is the hidden bonanza that really made the open source movement take off like it did. As it worked out, not that much software ended up in the open source ecosystem because some well meaning programmer decided that the world needed a better widget. Most of the software that ended up being open source was the result of work that was done to achieve a particular self-interested purpose, and would either “go to waste” if it weren’t shared, or the author preferred to share the burden of maintaining and advancing the software with others. Why share it? Because it’s less work that way!

It turned out that there were some very real financial incentives for individuals and organizations to participate in software sharing. That’s why they did it. It wasn’t because they hated Capitalism. It was because of Capitalism. They could save money on the front side by using free software instead of licensing commercial software, then when they had to do a little tweaking, or even had to write applications to run on top of the free software they’d used, it was to their benefit to share that code with other people in the community, because by collaborating on software they all needed instead of each person re-inventing the wheel in isolation, they could all save time and effort.

Now, this is the standard open source sales pitch, and to be perfectly honest, I’m sure there are many examples of companies that have gone down this route with the intention of saving money that could actually have spent less money licensing off-the-shelf software instead, when all is said and done. Some of these firms may even have been influenced by engineers who were more enthusiastic about using the open source software because of some ideological ulterior motive or even a simple hatred of Microsoft (or Oracle, or whoever). Open source is no panacea, but it’s not snake oil either. For the vast majority of open source software users, it’s all about money. And most of the largest implementations were only made after the green eyeshade folks gave it an enthusiastic thumbs up.

But we’re getting ahead of ourselves. At this point in our history of open source, it’s just making its transition from an intellectual exercise among a relatively tight group of academics, hobbyists, and corporate researchers into a growing phenomenon that’s being installed in corporations under the radar. It’s still small potatoes. No threat to the established commercial software companies. It’s the mid nineties.

Then the internet came onto the scene. Like the music industry, the software industry had been as much about promotion and distribution as production, if not moreso. Suddenly, software could be promoted and distributed at almost no cost, and the carefully-built system of salesmen, distributors, Value Added Resellers, and the like had been bypassed almost overnight. Luckily for them, the software industry had experienced quite a bit of consolidation and vendor lock-in by then, so the average business and home user was somewhat tied into the old system of Windows/Mac PCs running packaged software sold through old-style sales networks and retail stores. But even in the early days of the internet, cracks were already starting to show.

The problem was, the bread and butter of the software industry, medium-to-large businesses, had been increasingly resentful of being the targets of the software industry’s relentless profit maximization. Businesses are in the business of making money. And they make money by keeping revenues high and expenses low. By the 1990s, businesses were spending a staggeringly large amount of money on software, and due in large part to the managerial class’ emphasis of “features that sell” over “software that works well,” a large part of that investment was being wasted in ambitious software projects that ultimately failed. They were getting locked-in to vendors’ platforms and run on a costly forced upgrade cycle. Why did they do it? Because the alternative was writing and maintaining their own software in-house. Now, many, many firms did just that. But it was a headache and a risk. A Fortune 500 company can not take the risk that its lead programmer gets hit by a bus and nobody left alive understands how to keep its accounting system running. So many firms were locked into commercial software that they were not happy with.

So by the mid 90s, the fruits of this largely academic collaboration were starting to leak out over the internet: xBSD, Apache, Linux, Sendmail, BIND, Perl, MySQL, and many others. And it wasn’t just other academics and software hackers who were picking them up. It was corporations, large and small. The bedrock customers of the software industry. Was it idealism that brought open source software in? Partially, perhaps. Most open source software initially entered these firms under the radar: a Linux/Samba file server here, a FreeBSD/Apache web server there, built on old PCs from the junk room, usually. But when managers found out they’d been happily using free software for months, and in doing so had saved thousands of dollars, it got their attention. The IT managers and engineers at these firms liked that they didn’t have to go hat in had begging for funds to buy new software, and sometimes it saved them a lot of time that they would have spent reinventing the wheel. Sometimes, open source software made them look like heroes. Middle managers liked that they could roll out new software-heavy projects without having to beg the CFO for money. It made them look like heroes too. Upper management liked that they now had the leverage they needed to turn the screws on the software vendors who had been screwing them for so long.

Now a lot of this was bad news for the software industry. Any business that wakes up to find a strong new competitor in its market is unhappy. The owner of a main street five and dime who drives by a vacant lot outside of town and sees a “Coming soon: Wal-Mart” sign is a heck of a lot more threatened than Microsoft or Oracle is by Linux or MySQL. But there are some companies that have already been mortally wounded or even killed off by the availability of open source software in their niche, and others that are facing decreased prospects. If open source software continues in its ascent, the software industry will undoubtedly be transformed, and we might see a big drop in the kinds of profits that software companies have enjoyed over the past few decades.

The software industry is one of the United States’ most important industries. According to the BSA, the software industry makes a greater contribution to the US GDP than any other manufacturing industry. (Manufacturing makes up almost 14% of the US GDP, according to the US Bureau of Economic Analysis). More than 800,000 people are employed by the software industry, and they make an average of $69K per year (BSA). And this is more or less the case with most developed countries. Software makes up an important part of the world economy. How would the widespread adoption of open source software affect this segment of the economy?

It’s impossible to speculate with complete accuracy, but I think we could all agree on some basic generalizations:

It would open up the possibility of a shift of dominance away from the United States. Though much open source software originated in the US, there are no artificial controls to prevent people in other countries using and improving upon that software, and in fact obtaining de-facto dominance of any particular niche.

Some companies’ product lines are sure to suffer, and under-diversified companies might go out of business altogether. Intel and Linux delivered a 1-2 punch to companies like SGI and Sun; Oracle may face reduced profits in the future as open source databases like Postgres and MySQL attack their low-end market and creep up; Microsoft has already seen Linux and Apache prevent it from easy domination in the low-to-mid-end web server market.

It’s possible that we could see a fundamental shift in the software industry away from predominantly earning money from licensing fees into making money from support contracts, automated update and maintenance services, and consulting services. Some major software companies, like IBM, Oracle, and SAP already make a large proportion, if not the majority of their software-related money from services and support contracts, and that has been the case long before open source came on the scene.

If it ever got to the point that software users came to expect that most software would be free of charge, it would become difficult for an individual or small company to make any money by creating and licensing software. This is already shown itself to be true in the Linux subculture. So much of the software for Linux is free that there isn’t much of a market in shareware for Linux, as there is for the Windows and Mac platforms. FTP clients, utilities and other small, purpose specific software that would cost $10-20 (shareware) on Windows is generally available free of charge for Linux, and generally included outright on your typical Linux distribution. Individuals or small firms that might have an interest in distributing a shareware app on Linux probably just don’t bother.

So it’s likely that if open source software became more widespread there would be some negative impact on the economy. Some companies that are currently in business might be forced to change focus, survive with slimmer profit margins, or even go under as a result. Some companies that might have existed otherwise will never come to be at all. And individual countries, like the United States, might see their dominance in the software industry wane as the market is opened up to other, probably poorer, countries, like India and China.

So in the U.S., software company profits are likely to decline, and some jobs will probably be lost. There will be a negative economic effect from open source software.

Plenty of people have made similar claims, with the most vulnerable of the software firms, industry associations, and their paid mouthpieces being the most vocal. And they do have a point. The problem is, as is the case with any fact that’s promoted by an aggrieved party, that there’s another side to the story that’s conveniently ignored.

The side of the story that’s ignored is that there are two sides to the software industry: the producers and the consumers. In fact, the number of companies and individuals who use software vastly outnumber those that produce it. And there are two sides to corporate profit: revenues and expenses. The average corporation spends a huge amount of money, a significant percentage of its IT budget, on software licensing.

Take a look at the largest companies in the world. Of them, only Microsoft makes the majority of its money from software licensing. A handful of them, IBM, HP, Siemens, Hitachi, Motorola, Lockheed, Intel, make some of their money from software sales, generally a small part. A few, like Wal-Mart and Best Buy make a small portion of their revenue from retail sales of software. So the vast majority of large global companies consume software rather than produce it. Same goes for small businesses. And let’s not forget governments and academic and research institutions, some of the other largest consumers of software. When looked at it from this angle, if the cost of software is driven down by competition from open source, and thus a major cost of doing business is reduced for global industry, will it be a net gain or net loss to the economy?

And the cost savings can come in unexpected places. A large corporation that I know of replaced thousands of Sun/Solaris servers with inexpensive Intel/Linux ones and ended up saving $250,000 per year — in electricity — on top of millions of dollars in licensing fees. Where did that money go? Did it disappear into the ether? Well, Sun Microsystems (and the power company) probably thinks it did, but in fact, instead of going to Sun, it went to pay salaries, to fund new projects, and back to the shareholders in dividends and a kick in the stock price due to greater profits. In other words, it went right into the economy. Every dollar not spent on software licenses is spent on something else.

And economics is not a zero-sum game. Money can be spent in ways in which its positive impact on the economy is greater or lesser. If a firm spends $25 million developing a software product that never achieves widespread use and never makes much of an impact (and a huge proportion of commercial software projects fall into this category), the only positive impact on the economy will be the transfer of funds from company coffers to general circulation (and taxes) via the salaries of the employees involved.

On the other hand, if a useful piece of software becomes available at little or no cost to many companies, especially to companies that otherwise would not have been able to afford such software, it can give a major boost to that company’s productivity. In that case, even if no money was spent, all those companies increased their efficiency and revenues. Increased productivity and decreased expenses can have a massive effect on a company’s bottom line.

So let’s look at that quote again:

I guess free software foundations are going to employ people from now on. Its the same evil mega corporations that employ hundreds of thousands of people and make the world economy function. Make them “smaller, weaker, and easier to keep in their place” and raise the unemployment rate to double digits not to mention lowering the standard of living world wide I suggest voting NO for RMS Democracy.

Is it “free software foundations” that pay the paychecks of most of the people developing open source software? No, for the most part it’s for-profit enterprises (and some schools) who have a self-interest in producing the software. And who is it who’s using all this software? Long-haired hippies running servers for anti-capitalistic websites? Nope, it’s everyday IT folks doing everyday work for regular companies.

Open source software has been a tremendous boon for the thousands of small and medium sized businesses that have been mostly shut out of the enterprise software markets, both as producers and consumers. A few small companies like Red Hat and SuSE have been very successful in producing and supporting open source software, and countless small consulting firms have found open source to be a revelation.

Let’s have an example: A small consulting firm is hired to solve a problem. The client has a budget of $5000. They could solve that problem in 50 hours by coding it from scratch (earning $100 per hour), or in 20 hours using pre-existing open source software (earning $250/hr). In many cases like this, even if there is commercial software, it would cost more than $5000 just for the license. So this small firm rolls out the open source software, with modifications, makes its $5000, then turns around and sells a similar system to its next client, making another $5000. And the clients get the additional peace of mind knowing that even if this small consulting firm goes away, since their system is built on a known platform, someone else should be able to pick up the pieces. In this case, everybody wins. In fact, the commercial software firm doesn’t even lose, because they wouldn’t have been interested in this business anyway.

The established interests (software firms and their paid consultants) have been playing this sob story of how open source is putting them out of business and how it’s a threat to the economy, insecure, more expensive, etc, etc. They’ve even been lobbying the government. It’s the same old sob story we’ve heard over and over, and apparently there are plenty of people who will believe it.

I just hope that they’re not successful in implementing any kind of artificial protections for themselves through lobbying, and I say this as a proponent of free trade. Protectionist tariffs and subsidies on domestic agriculture, raw materials, and manufactured goods are a constant temptation for a country that encounters competition from outside. Europe and the United States to this day continue to artificially prop-up domestic industry even as their governments preach the benefits of free trade and open markets. Who are the ones that want to open the markets? Ironically, it’s sometimes the same folks who would call open source software proponents communists.

The outcry over open source’s negative effect on the economy is similar to the U.S. steel situation. The U.S. steel industry is having a hard time competing with imports from abroad, so they successfully lobby the government into slapping tariffs on foreign steel. That keeps steel prices artificially high, and subsequently the U.S. automobile and construction industries suffer. A few steel workers keep their jobs, a few auto workers lose theirs. I’m not familiar enough with the situation to know whether it was a net gain or loss for the country’s economy, but the point is that you take away from one side, you give to another. That’s the way things work.

Here’s another example: Wal-mart has been bad news for small businesses all over the USA, and for their owners and employees. It’s been bad for the downtown business districts, some of which have been decimated, while others have been merely forced to transform into restaurant and entertainment promenades. But in the larger picture, has it been bad for America? Well, ask the millions of rural Americans who now have access to socks for $2 per dozen and $35 DVD players. They’ll say it’s not so bad to have some economy of scale leveraging their purchasing power. (Disclaimer: personally, I hate Wal-Mart, and I love vibrant downtown business districts and small, quirky businesses.) But the point is this: just as we might decry the negative impact that foreign trade or big box stores might have one segment of the economy or society, the net result has been that these factors mean that a couple hundred million Americans can now buy a heck of a lot more for their money that they could twenty years ago.

In conclusion, the number of people who will be negatively affected by the availability of high quality, low cost software is relatively small. The software industry may need to transform, and some firms may not survive, but the overall impact on the economy will be positive. Just because the oil companies are enjoying increased profits due to higher gas prices does not mean that high oil prices is good for the economy. Quite the opposite.

We might have heard a similar story a couple hundred years ago, during the industrial revolution: “But millions of people are employed planting and harvesting, and the tractor will put them all out of business! And the cotton gin will put all those people picking the seeds out of cotton plants out of work! And mechanical looms will put all those weavers out of work! Oh the humanity!”

Every time there’s a transformation in one segment of the economy, we hear the same outcry. And while it may be downright tragic for the affected parties, the rest of us can’t get too caught up in the drama.

History is chock full of well-meaning people succeeding with their plans and inadvertently making life worse for many others. The early Communists thought they were fighting for freedom. Freedom is a good idea, but if you go about promoting it in the wrong way, you can end up making people less free. You may not notice what’s happening until it’s too late because ideology can blind you. Many people today who are fighting for intellectual property rules because they think it promotes innovation and progress may actually be actually hammering nails into innovation’s coffin.