Repeal the Microsoft Tax

Now that the Justice Department and Judge Colleen Kollar-Kotelly are content with the wrist-slapping meted out to Microsoft, few expect any abatement in the company’s abuse of its monopoly power. Although many argue that, eventually, the markets themselves will bring about more balanced competition, the markets that Microsoft dominates operate on different economic principles than most others. The last twenty years brought such dramatic technological change that it’ll be much harder for competitors to dislodge Microsoft from its perch atop the industry than it was for Microsoft to dethrone IBM a generation ago.

Editorial notice: All opinions are those of the author and not necessarily those of

When IBM dominated the industry, computers were used primarily for performing calculations and automated tasks. But as Microsoft rose to prominence, computers were becoming used more and more to communicate. Although such a change might seem insignificant, it caused a radical shift in the economic dynamics of the industry. Instead of operating under the traditional rules of supply-and-demand, where scarcity increases value, the industry began to operate according to the principles of network economics, where ubiquity increases value.

Networking Made Simple

Network economics governs any technology where the ubiquity of the technology directly affects its utility. This is typically the case with communication technology; as the technology becomes more widely deployed, more options for communication are created, thereby increasing the usefulness of the technology. The market for fax machines follows this principle: the utility of a fax machine is proportional to the number of other fax machines with which it can communicate.

The very first fax machine sold was essentially useless, because it couldn’t communicate with anything. The sale of the second fax machine actually increased the utility of the first by creating a potential communication channel that didn’t exist previously. Whenever a new fax machine is put into use, it increases the potential of every other fax machine to communicate. Therefore, each fax machine sold increases the utility of fax machines in general.

Obviously, this is quite different from the traditional economic dynamics that govern most products. For example, if you buy a jacket, whether or not anyone else already owns that type of jacket does not change the jacket’s usefulness to you. Fashion considerations aside, whether you’re the only one who owns the jacket or you’re one of a million who own it, the jacket will still function exactly the same and keep you just as warm.

A Very Brief History of Modern Computing

During the decades that IBM reigned over the computer industry, most computers were huge machines in complex data centers at universities, large corporations, and government agencies. Computer networking was in its infancy, so the most common way to share data was by physically transporting cumbersome storage media. Very rarely was data shared across the bounds of an organization. Most computers were islands unto themselves.

But by the mid-1980s–coincidentally when IBM began to surrender its position to Microsoft–this began to change. The computing universe no longer consisted of a few installations in large organizations; most businesses now owned personal computers. Nor were computers any longer the exclusive domain of technical specialists. Many, many more people were creating data, and that data frequently needed to be shared. At first, people exchanged files on disks, which by then could fit in a pocket. Small on-line services called bulletin board systems sprang up, bridging distance by allowing people with modems to form communities. And when the Internet came into widespread use in the early 1990s, the masses discovered that sharing files could be as easily as sending e-mail.

The explosive rise in the volume of data being shared meant that computers were becoming increasingly used as tools for communication. As a result, the industry’s market dynamics began to resemble those of fax machines. Computer technology was now subject to network economics.

Protocols and Standards

Whenever information needs to be shared among machines, those machines must agree on the format of that information. After all, if you send me a file that I can’t open, you might as well not have sent me the file at all. To be useful, communication must be understood by all parties involved. In technology, useful communication is enabled by common protocols, or rules that specify how communication takes place.

Protocols come in two flavors: open and proprietary. Open protocols are published specifications that may be adopted by any vendor. Proprietary protocols, on the other hand, are usually owned by a single company that views them as strategic assets to be guarded from competitors.

At some critical mass, a protocol becomes a standard if it is widely adopted enough. A given protocol can be considered standard once it becomes the default choice in the marketplace.

When Protocols Become Standards (Or Not)

Markets for communications systems develop differently depending on whether a given protocol becomes standard and whether that standard is open or proprietary. Generally, one of three scenarios will play out:

* The first possibility is that no protocol–open or proprietary–reaches critical mass in the marketplace. Let’s say there are five vendors of fax machines, all of them using incompatible protocols, and each vendor captures 20% of the market. Any fax machine purchased is therefore limited to communicating with only 20% of the fax machines in use. The overall utility of the fax machine is limited, which depresses demand for them. In the absence of a standard, the market stagnates.

* Another possibility is that an open protocol becomes standard, which is what occurred in the real-world case of fax machines. Unlike the first scenario, the market is not hampered by fragmentation. The network will grow much larger, leading to greater utility from fax machines. And, because no one company controls the protocol, fax machines are commoditized, creating price competition among vendors. This scenario provides the greatest utility at the least expense to the customer, but is the least desirable to vendors harboring monopolistic ambitions.

* In the third scenario, a proprietary protocol achieves market dominance. For example, let’s say one company develops a fax machine incompatible with all others, and the company seizes a large share of the market through excellent marketing. In this case, fax machines from that vendor would provide the greatest utility, while the rest–unable to communicate with those from the dominant vendor–would be virtually useless. Customers wanting the functionality of a fax machine would have little choice but to purchase one from the market leader. A monopoly develops, solidified by switching costs: ditching the dominant vendor requires replacing most of the machines in use. Once a proprietary protocol becomes standard, monopolies can be built fairly easily and
defended with little effort.

Microsoft’s Document Monopoly

Microsoft Office currently captures most of the market for office suite software. Office suites are collections of commonly-used applications essential to any organization; they typically include word processors, spreadsheets, databases, and graphics programs.

Every day, hundreds of thousands of documents–things like resumes, contracts, presentations, sales data analyses, etc.–are created using Microsoft Office. People create most of these documents with the intention of sharing them with others; very rarely do they stay exclusively on the computer used to create them.

Because Microsoft Office documents are used so frequently for data sharing, they have effectively become communications protocols. And, because Microsoft controls the file formats of these documents, the company controls a vital standard, allowing it to exercise and defend its monopoly with very little possibility of being challenged by competitors.

The use of Microsoft Office documents is so pervasive that they’re essential to participation in the modern economy. Although there are programs from other sources capable of interacting with Microsoft Office documents, none of them are 100% compatible; owning a copy of Microsoft Office is the only way for a business to guarantee that it can communicate fully with others. As long as these documents continue to be the lingua franca of data exchange for business, network economics will protect Microsoft’s monopoly in the office suite category.


Monopolies are aberrations that perpetuate market inefficiency. Nobody benefits from a monopoly but the monopolist; everyone else is harmed by stifled competition. In an efficient market, when a company starts acting in opposition to its customers, those customers can patronize other vendors. This helps align the interests of buyers and sellers. But monopolists can ignore or even work against the interests of customers with little fear of losing them; they’ve got nowhere else to go.

A fundamental element of free markets is the customer’s freedom to choose. In the case of Microsoft, customers are “free” to choose alternatives, but to do so, they must sacrifice the ability to communicate. Is it reasonable to expect customers to make that sacrifice? Not when you’re talking about a technology whose primary purpose is to enable communication.

A Solution

Conservatives tend to be philosophically opposed to government intervention in markets. Perhaps this explains the lack of enthusiasm that the Bush Administration displayed for the Microsoft antitrust case. (Although it doesn’t explain why the Bush Administration favored steel tariffs or farm subsidies.) But, if the argument for action were instead framed in terms of cutting government procurement expenditures, if a solution were found that weakened Microsoft’s monopoly through the markets rather than direct government intervention, and if the solution applied to the industry as a whole and not Microsoft specifically, the Bush Administration might be much less likely to stand in the way.

The federal government has tremendous buying power, as do many state governments. This buying power can be harnessed to encourage open standards in markets driven by network economics. Specifically, Congress should pass–and the President sign–a law mandating that all commercial software applications purchased by the government use open protocols for file formats and data communication. State governments should consider adopting similar laws.

Of course, such laws would need to be written to prevent companies like Microsoft from pulling their usual “embrace and extend” tricks, which are merely intended to pollute open standards and give the monopolist control of them. A transitional period would also be required to allow vendors time to document their protocols and submit them to accepted industry standards bodies, which would assume responsibility for their future development.

This solution would loosen Microsoft’s grip on crucial industry infrastructure. It would also bring greater competition to other segments of the software industry, because Microsoft now uses its monopoly in business applications to fortify its other monopolies. Microsoft’s decision against providing a version of Office for the Linux operating system, for example, helps ensure that Linux isn’t more widely adopted by businesses.

Meet the New Tax Man

Microsoft’s artificially high licensing fees and unreasonably restrictive licensing terms are so legendary within the industry that their economic impact has become known as the Microsoft tax. (Did you know, for example, that the cost of almost every PC sold includes the cost of a Microsoft Windows license, even if the purchaser intends only to use a different operating system on the machine? Or that Microsoft is scheming to have universities force every student to pay for Microsoft software, regardless of whether they actually use it?)

Microsoft’s monopoly gives it the power to collect fees far in excess of what a truly free market would bear. Microsoft is able to extract this additional money from customers–in some cases, unwilling and even unwitting customers–simply because it is not subject to genuine competition.

Because customers have little choice but to render unto Microsoft whatever it demands, the Microsoft tax, like any tax, effectively diverts capital from what would have been its natural, market-based flow through the economy. This diversion of capital represents a market inefficiency that, if corrected, would free up capital for other uses. With the economy in its current state, with businesses starved for capital that could get our economy growing again, would finding ways to repeal the Microsoft tax really be such a bad thing?

Copyright 2003, Evan Coyne Maloney

About The Author:
Evan Coyne Maloney is a political commentator and software developer living in New York City. Some of his writings can be found on the website Brain Terminal, at:


  1. 2003-01-23 7:39 am
  2. 2003-01-23 7:41 am
  3. 2003-01-23 7:44 am
  4. 2003-01-23 7:45 am
  5. 2003-01-23 7:56 am
  6. 2003-01-23 8:07 am
  7. 2003-01-23 8:12 am
  8. 2003-01-23 8:14 am
  9. 2003-01-23 8:20 am
  10. 2003-01-23 8:32 am
  11. 2003-01-23 8:33 am
  12. 2003-01-23 8:41 am
  13. 2003-01-23 8:42 am
  14. 2003-01-23 8:46 am
  15. 2003-01-23 8:48 am
  16. 2003-01-23 8:50 am
  17. 2003-01-23 9:57 am
  18. 2003-01-23 10:48 am
  19. 2003-01-23 11:06 am
  20. 2003-01-23 11:32 am
  21. 2003-01-23 11:39 am
  22. 2003-01-23 12:10 pm
  23. 2003-01-23 12:39 pm
  24. 2003-01-23 12:43 pm
  25. 2003-01-23 1:21 pm
  26. 2003-01-23 2:04 pm
  27. 2003-01-23 2:29 pm
  28. 2003-01-23 2:32 pm
  29. 2003-01-23 2:54 pm
  30. 2003-01-23 2:58 pm
  31. 2003-01-23 3:18 pm
  32. 2003-01-23 3:24 pm
  33. 2003-01-23 3:29 pm
  34. 2003-01-23 3:41 pm
  35. 2003-01-23 3:46 pm
  36. 2003-01-23 3:59 pm
  37. 2003-01-23 4:28 pm
  38. 2003-01-23 4:30 pm
  39. 2003-01-23 4:36 pm
  40. 2003-01-23 4:56 pm
  41. 2003-01-23 5:04 pm
  42. 2003-01-23 5:28 pm
  43. 2003-01-23 6:26 pm
  44. 2003-01-23 6:35 pm
  45. 2003-01-23 6:42 pm
  46. 2003-01-23 6:49 pm
  47. 2003-01-23 6:52 pm
  48. 2003-01-23 7:04 pm
  49. 2003-01-23 7:20 pm
  50. 2003-01-23 7:35 pm
  51. 2003-01-23 7:39 pm
  52. 2003-01-23 7:54 pm
  53. 2003-01-23 10:56 pm
  54. 2003-01-23 11:47 pm
  55. 2003-01-24 12:54 am
  56. 2003-01-24 1:36 am
  57. 2003-01-24 3:07 am
  58. 2003-01-24 3:10 am
  59. 2003-01-24 3:28 am
  60. 2003-01-24 4:29 am
  61. 2003-01-24 2:08 pm
  62. 2003-01-24 2:19 pm
  63. 2003-02-01 8:35 pm