Note: All footnotes, which contain a great deal of detailed information, are excluded from this web version of the paper.
In 1986 WordPerfect and WordStar were the two leading word processors, with moderately larger market shares than several other programs (DisplayWrite, Word, MultiMate, Samna Word). In the next few years, however, WordPerfect broke away from the pack, and by 1990 it clearly dominated the market. This history is tracked in Figure 8.8, which portrays market shares of the leading word processors.
In the late 1980s Microsoft shifted its attention away from its DOS version of Word, which was then number two in the market, and began to focus on the Windows version. Microsoft Word for Windows hit the market in late 1989, and began to generate a serious market share in 1990. By the time Windows became the dominant operating system, the two leading Windows word processors were Ami Professional and Microsoft Word.
After the introduction of Windows 3.0, Microsoft Word grew at an extraordinary rate. By 1994 it dominated the market more completely than WordPerfect had, even during its peak years. WordPerfect for Windows, which was introduced in late 1991, a year after its OS/2 product was introduced and almost two years after Windows 3.0 was introduced, jumped initially to a 20% market share, but then its market share stabilized and began to decline. Ami Pro showed some good growth (in terms of units) but never achieved much of a market share in terms of revenues.
WordStar, the original leader in the word processor market, lost its market position by failing to keep its quality at the level of its competition. The story was different with WordPerfect. Unlike WordStar, WordPerfect was always judged a high-quality product in its DOS incarnation. Where WordPerfect fell behind was in developing a high-quality product for Windows, and in incorporating its product into a high-quality office suite.
The dominance of DOS WordPerfect is clear in the reader ratings of PC World Magazine (Figure 8.9). And the story in the reader ratings accurately reflects the history of market share. In 1984 WordStar was the leading product in the reader ratings, followed by Microsoft Word (DOS) and MultiMate. WordPerfect was not yet a market leader. By 1987, however, WordPerfect was in first place. Its ascent continued until late 1991. During the 1990s, however, we see a decline in DOS word processors and a rise in Windows word processors—Word for Windows, WordPerfect for Windows, and Ami Pro for Windows. In the Windows market, Word won out.
In short, with the change in operating system came a change in leadership. And once again, measures of quality explain that change. Figure 8.10 shows the number of wins in the DOS word processor market for various products. The closest competitor to WordPerfect DOS is Microsoft Word for DOS, but after 1989 WordPerfect is just about the only game in town when it comes to quality DOS word processors. A similar story is told by ratings of different DOS word processors over time (Figure 8.11). The advantage of WordPerfect over Microsoft Word is, however, somewhat less discernible in this chart.
Thus, the preponderance of evidence indicates that WordPerfect deserved its leading market share in the DOS market. The scale of WordPerfect’s domination may be a little surprising, given that Microsoft Word ran such a close second, but the poor showing of the number two product is a common thread throughout these chapters. This could be the result of instant scalability, or perhaps network effects and economies of scale.
WordPerfect did have an additional advantage (left out of our analysis) of offering near-legendary technical support, but in this, after all, the company didn’t have much choice. WordPerfect for DOS was notoriously difficult to learn, as anyone who remembers the keyboard templates for the product can attest. The program was run by using arcane combinations of the twelve function keys with the control, alt, and shift keys. Technical support was more important for such a product than for the more intuitive Windows products.
As an operating system, DOS had about run its course, but the change to a new operating system was not going to come easily. Programming for Windows required a whole new set of skills. This is evidenced by the number of producers of DOS word processors that tried to buy Windows products instead of developing their own:
Microsoft Word ultimately came to dominate the Windows world for pretty much the same reasons that Excel succeeded: Microsoft Word was a superior product at a time when consumers were rethinking their adoptions, and Microsoft engaged in better strategic marketing (office suites).
The quality advantage enjoyed by Microsoft Word is documented in the magazine ratings. Figure 8.12 shows the relative ratings of the three leading Windows word processors. Microsoft Word almost always won these magazine comparisons. Word’s closest rival was Ami Pro, which started with a much smaller market share and installed base. WordPerfect was late out of the gate with its Windows product, and when it arrived it was poorly reviewed:
The "too much WordPerfect" criticism won’t bother the DOS word processor’s many fans, but believe me, folks, it keeps the program from being a great Windows application. The menus are bizarrely organized. …These things are the price the program pays for being faithful to the DOS version
Moreover, the program’s rawness compared to Ami Professional 2.0 and a forthcoming version of Word for Windows points out a greater shortcoming. Ami Pro 2.0 is the fourth or fifth Windows word processor that its design team has created. While the new Word for Windows (expected by the end of the year) is only the second Windows word processor created by its team, the company has also learned from the highly successful Macintosh version of Word. In contrast, this is the first Windows product designed by WordPerfect Corporation, which has not scored well with its GUI word processors on other platforms.
[WordPerfect for Windows] first release suffers from bugs and UAEs that will probably disappear from an interim release early in 1992. …WordPerfect for Windows doesn’t feel fully at home in its GUI environment. ... Its limitations derive from the DOS version and won’t bother users of the DOS version.
The large installed base of WordPerfect customers had their allegiance tested first by WordPerfect’s late entry into this market, and second by the poor performance of WordPerfect when it did enter. WordPerfect didn’t produce a reasonable Windows product until the end of 1993. This delay set the stage for WordPerfect’s fall.
If the Windows versions of word-processing programs had shared characteristics with earlier DOS versions from the same vendor—or at least had greater ability to import the older DOS files—this backward compatibility might have enticed owners of the DOS products to upgrade to the same vendors’ Windows products. But the producers of many leading DOS word processors were not able to produce their own Windows products; they simply purchased Windows products from outside providers and attached their names. In the process, they sacrificed backward compatibility, which certainly affected the purchase decisions of their former customers who were upgrading to Windows.
Sacrificing sales to former customers was not always a bad idea. Consider, for example, the history of Ami Pro. The installed base of SamnaWord customers was very small—a clear hindrance to its success if it were to remain tied to its installed base. It would have made sense to the owners of Ami to design the best product they could and ignore the consequences to their installed based. As it was, Ami Pro (for Windows) achieved a far greater market share than its previous DOS incarnations.
Of the producers of Windows word processors, only the creators of Ami Pro and Word had experience in producing word processors that fit well into GUI environments. Thus, it should come as no surprise that these two products dominated the magazine reviews.
Figure 8.13 illustrates the number of wins for Windows word processors, which, is almost completely dominated by Word for Windows and Ami Pro, as was the case in Figure 8.12.
The advantages of Word are less apparent in Figure 8.13 than in figure 8.12, however. In fact, the relative quality between MS Word and Ami Pro is very similar to the relative quality of MS Word for DOS and WordPerfect for DOS in the late 1980s. But even if we were to conclude that Word and Ami Pro were exactly equal in quality, which is inconsistent with the results in Figure 8.12, there are several reasons why Word would be expected to do better in the marketplace than Ami Pro. First, as already noted, Microsoft Word for DOS had a much larger installed base. Second, Ami Pro went through several owners (Samna, Lotus, IBM) in a fairly short period of time. Third, Microsoft had its office suite on the market first, at a lower price and, as we have seen, Excel was a more powerful companion spreadsheet than 1-2-3.
As in the spreadsheet market, price differences among the major programs did not play a very important role in the evolution of the word processor market. Price was used strategically, but competitors ordinarily matched each other’s prices.
Figure 8.14 shows prices (received by vendors) for the major word processors. As in the spreadsheet market, the most important trend is the decline in prices that began when Windows 3.0 was introduced in 1990. The resulting competitive jockeying led to price wars. It is important to note, however, that there is no evidence of price declines until Microsoft Word began its assault on WordPerfect's position.
In 1990, 1991, and 1992 Ami Pro was more expensive than Word for Windows. These were important transition years, yet Lotus provided no economic incentive for users to pick Ami Pro over Microsoft Word. Both were ranked higher than WordPerfect, but Word had a slight edge in quality, a larger installed base, a lower price, and was part of a superior office suite. Ami Pro did not lower its prices until Lotus was purchased by IBM. By the time Ami Pro became price competitive (1993), Microsoft had 40 percent of the market and the reviews were about to start concluding that Ami Pro was no longer competitive with Microsoft Word.
As for WordPerfect, the price for its late and inferior Windows product was virtually identical to that of Microsoft Word through 1995. After that, it was priced below Word, but the price cut was a case of too little, too late.
The transition from DOS to Windows presented a golden opportunity for ratcheting up market share in the word processor market, and Microsoft was positioned perfectly to take advantage of this opportunity. Microsoft offered a Windows-based word processor that was clearly better than incumbent market leader and as good or better than any other challengers. Microsoft Word for Windows was a much better product than WordPerfect for Windows, and it was available much sooner. Ami Pro was close in quality to Word, but Microsoft had a larger installed base and a superior and more stable market plan. Furthermore, Microsoft had a top-flight spreadsheet, Excel, which it was able to bundle with Word into a very attractive office suite. This combination of factors proved irresistible.
In each market, the highest-ranked product racked up the largest market share. Whether the shift from WordPerfect to Microsoft Word was "too" slow, "too" fast, or just right, we don’t know. Certainly the market-share changes in the word processor market—first WordPerfect’s growth, and then its replacement by Microsoft Word, all within a decade—are more rapid than we find in many markets. The history of the word processor market certainly doesn’t provide any support for the claim that that a market will standardize on an inferior product, or that consumers lock in to products.
Another important feature of this market is the smoothness of the changes in market share. There are no sudden accelerations or decelerations that might be consistent with tipping. After WordPerfect had grown to almost 50 percent of the market, a position that might have been thought consistent with the market having tipped in its favor, it actually began to lose market share in favor of a product with a much smaller market share. Once again, the empirical evidence fails to support the theoretical construct of tipping.
We have decided to group the spreadsheet and wordprocessor analysis for the Macintosh market into one section because of certain similarities and comparisons to one another that we wish to make.
In both markets, the Microsoft product was the market leader as well as the highest quality product.
In the spreadsheet market, Microsoft Excel was clearly the leading spreadsheet in terms of quality. Figure 8.15 indicates that Excel won every numerical review that we found. For all reviews that rated products, Excel won in 10 of 12 instances. In MacUser Magazine’s ratings of spreadsheets Microsoft Excel scored an average of 4.79, Lotus 1-2-3 averaged 4.2, Resolve scored 4.5 and Full Impact scored 3.33 (out of 5 points). The small number of reviews might be a cause for some concern, but this small number may be due in part to the fact that this market settled early on a single product—Microsoft Excel.
The degree to which sales of Excel have dominated the Macintosh spreadsheet market is quite remarkable. In fact, it would be fair to say that it is rare to find a market where one product is as dominant as Excel has been in the Macintosh spreadsheet market since 1988. Figure 8.16 displays the market shares of Excel and several other spreadsheets. One has to look hard even to see the market shares of Excel’s "major" competitors on this diagram.
A similar story can be told for the wordprocessor market. Microsoft Word was the leading word processor for the Mac from the time of our earliest data (1988). Alternative products included MacWrite, WordPerfect, WriteNow, Full Write, and Nisus Writer. Figure 8.17 shows trends in the market shares of the leading products.
Although the number of available reviews is small, Word won 59 percent of them (10 of 17). In MacUser Magazine’s ratings of word processors over several years Microsoft Word scored an average of 4.21, MacWrite scored 4.0, WordPerfect scored 3.7 and Nisus Writer scored 3.88 (out of 5 points). Word was clearly the highest rated word processor in the Macintosh market.
We can conclude, therefore, that Microsoft’s two entries in this market, Word and Excel, were the best of breed.
Some particularly revealing evidence has to do with a comparison of Excel and Word in the PC and Macintosh market. First, note that Microsoft achieved dominance in the Macintosh market considerably earlier than in the PC market. This is clearly seen in figure 8.18.
Examination of Figure 8.18 reveals that Microsoft achieved very high market shares in the Macintosh market even while it was still struggling in the PC market. On average, Microsoft’s market share was about forty to sixty percentage points higher in the Macintosh market than in the PC market in the 1988-1990 period. It wasn’t until 1996 that Microsoft was able to equal in the PC market its success in the Macintosh market. These facts can be used to discredit a claim sometime heard that Microsoft only achieved success in applications because it owned the operating system, since Apple, not Microsoft, owned the Macintosh operating system and Microsoft actually competed with Apple products in these markets
Our main interest here, however, is in comparing the Macintosh price to the PC price for the same product sold in two markets, one market where the seller has a dominant market share and another market where the seller’s market share is relatively small. This will allow us to test the structuralist theory of monopoly and whether Microsoft acted like a monopoly.
Economists have historically been of two views with regard to the relationship between market share and monopoly behavior. Market structuralists, following in the footsteps of Joe Bain, argue that high market share almost always leads to monopoly behavior. Critics of this theory, collectively known as the Chicago School, disagree. They argue that even if one firm dominates a market, in the absence of barriers to entry, potential competition will cause the market to behave competitively.
The behavior of Microsoft in these markets supports the Chicago School view and also supports Microsoft’s contention that it has neither harmed consumers nor acted like a monopolist.
Excel’s Macintosh market share of 70% in 1989 would indicate monopoly to a structuralist. In the PC market, on the other hand, Excel was just getting off the ground. In 1990, when Microsoft Word for Windows was first being launched, Microsoft Word for the Macintosh had a market share of almost 60%. Again, Word’s Macintosh market share would appear monopolistic to a structuralist.
Those holding a structural view of monopoly would expect Microsoft to charge a monopoly price in the Macintosh market and a competitive price in the PC market.
Price comparisons suggest the opposite of what a structuralist would expect. PC-Excel, far from being cheaper, in fact averages 13% higher prices than its Macintosh cousin during the period 1988-1992. PC Word’s price, on average, was more than eighty percent above the price for Macintosh Word prior to 1993. Both pairs of prices become virtually the same later in the decade. To double check this result we went to price advertisements in computer magazines. We used data from a single retailer selling to both markets (PC Connection and Mac Connection) and compared prices. Excel on the PC was consistently about 33% higher than for the Macintosh version. A similar result holds for Microsoft Word.
In other words, Microsoft did not act like a textbook monopolist (charging higher prices) in the market where it appeared to have a structural monopoly. In fact, it charged lower prices in the market where it was dominant than in the market where it was competing. Furthermore, as we saw earlier in the chapter, after Microsoft came to dominate the PC word processing and spreadsheet markets, it lowered prices again—dramatically.
What might be going on, then?
One answer is potential competition—i.e., Microsoft worries about competitors even when it has a very large market share. Potential competition is capable of causing a monopolist to behave like a competitive firm. In an industry where entrenched incumbents such as Lotus and WordPerfect were eclipsed in relatively rapid fashion, this might seem to be a very real concern. Microsoft’s concern about losing its markets might also explain why Microsoft has not lost any markets it has gained. The decline of Lotus and WordPerfect might have been due to an erroneous lack of such concern.
Falling prices, of course, don’t have to last. A firm might initially lower prices (as Microsoft did as it was conquering the PC spreadsheet market) as a predatory strategy intended to drive out competitors. The facts, however, do not support any rational predation story. If Microsoft were using predation to achieve victory in the PC market, it would be irrational to charge equivalently low prices, to say nothing of lower prices, in the Macintosh market. Furthermore, predation requires low prices now (to destroy a rival) and high prices later (after the rival is gone). But the pricing story here is prices that fall, and then fall further. It’s hard to make a case for predation in that.