Merrill Lynch: war will not hit IT spending

The war in Iraq won't slow IT spending, but a quick resolution of the war won't boost spending either, according to U.S. and European CIOs surveyed by Merrill Lynch.

Of the 100 respondents, 73 said the war in Iraq would not slow spending, while 17 said it would. A quick end to the war would only lead to a spike in spending with 10 of the respondents, while 90 said it would not, according, to the survey.

Investment bank Merrill Lynch released the results in a report late last week. Of the 100 CIOs surveyed, 75 work in the U.S. and 25 in Europe.

Even 17 percent of CIOs deciding to cut their spending would spell trouble for vendors with less recurring revenue, such as storage specialist EMC and Sun Microsystems, according to Merrill Lynch. The survey was conducted before the start of the war.

"In general, spending is hampered as much by structural problems in the economy as by war," Steven Milunovich, a vice president with Merrill Lynch, wrote in the report.

Away from the war, the survey found that 66 percent of the CIOs would like to increase the variable cost portion of their IT budgets, but 65 percent don't see the utility computing schemes being promoted by vendors as the way to do that.

Fixed costs such as facilities and staff make up about 70 percent of an IT budget today, and IBM, Sun, and Hewlett-Packard promote "pay as you go plans" to swing that ratio. However, CIOs say these won't be a real option until 2006, according to the survey. More evangelizing is needed for utility computing, Milunovich wrote.

Meanwhile, CIOs are spending more to buy servers. Of the CIOs surveyed, 55 percent said spending on Windows servers is increasing, 38 percent said more money is going to buy Unix servers and 33 percent said spending on Linux servers is up. Budgets for mainframes are down or flat, said 63 percent of the CIOs.

The interest in Unix servers is surprising, and suggests that Linux is not replacing Unix as quick as some in the market may have thought, Milunovich wrote. The continued decline of mainframes is not new and mainframe rebirth is not expected, he wrote.