Open question

The government claims open-source software means a 60% saving. It doesn't add up.

Israel's Ministry of Finance is currently exploring the possibility of abandoning commercial software, now in use in the public sector, in favor of "cheaper" open-source alternatives. Of course, it is a good thing that Ministry officials are looking into ways to reduce expenditures on software procurement, considering that the State's computerization budgets for 2004 have been cut by 25%. However, it is doubtful that switching to open-source software will reduce overall procurement costs. The total cost of ownership of open-source software is actually greater than the total cost of ownership of proprietary software. It is therefore very curious that Ministry officials claim that switching to open-source software products will save Israeli taxpayers up to 60% of the cost of continuing to do business with Microsoft and other proprietary software companies.

Not really cheaper

There is a common misconception that open-source software, such as Linux, is cost-effective because it can be freely loaded on as many computers as desired without incurring additional license fees. In fact, software license fees comprise only a small percentage of the total cost of ownership (TCO) of software. Most estimates place software acquisition costs between 5% and 8% of TCO. TCO is dominated by costs related to customizing systems, maintaining and servicing systems, and training systems personnel.

So what's wrong with saving money on licensing fees and 5%-8% of TCO with a switch to open-source software? The problem is that the costs of customization, maintenance and training are usually considerably higher under open-source solutions. A study published last year by the research firm IDC compared the TCO of an open-source Linux-based server to the TCO of Microsoft Windows 2000. The study found that the TCO of Windows was lower than the TCO of Linux in four out of five common workload scenarios. IDC surveyed 104 medium-sized and large enterprises operating in various industries in North America and concluded that, over a five-year period, Windows is 11 to 22.5 percent cheaper to own than Linux.

Although Microsoft Corporation commissioned the IDC study, it is generally accepted that IDC did not rig the results. IDC has a number of large corporate clients and it used the same methodology as in other cost-comparison studies that it has performed in the past. Another factor that adds credibility to the IDC results is that leading open source advocates accept the findings of the study. They acknowledge that open-source software cannot currently beat Microsoft on a TCO basis.

A particularly interesting aspect of the IDC study is that it finds that the bulk of the savings associated with operating Windows comes from lower personnel costs. Personnel costs constitute between 50 and 70 percent of TCO. This finding is especially relevant in the Israeli context. In light of the tendency to pay unjustifiably large salaries in the Israeli public sector, the increase in TCO that would accompany a switch in the Israeli public sector to open-source software could set new cost increase records. So how is it that the Ministry of Finance estimates a reduction in costs of 60%?

Insulation from market forces

According to well-known economists Josh Lerner and Jean Tirole, in an article recently published in the Journal of Industrial Economics, open-source software development needs to overcome a number of difficult problems before wide-scale adoption of open-source solutions in industry and government becomes feasible. Two of the more serious problems with open-source software are the "forking" of open-source projects and the orientation of open-source products towards high-end users.

The forking of open-source projects occurs when passionate disputes between open-source software developers over product design lead to the splintering of projects into a multitude of varieties. With proprietary software, forking generally does not take place since development is centralized within a firm and disciplined by market forces. Relying on a current open-source product design is, therefore, inherently risky.

The orientation of open-source products towards high-end users is also a result of open source software developers’ insulation from market forces and organization into disparate voluntary associations. Open source software developers virtually design products for their own use. Proprietary software firms, on the other hand, develop their software according to the needs of their most computer-challenged customers.

'Shared' rather than 'open'

There is no denying that open-source software development has, in the past, been a critical element of software innovation. Open-source software research in universities has often been an important first step in the successful commercialization of software products. As long as such research remains available under terms that do not limit its application in commercial products, university-based open-source research can be expected to continue spurring software innovation in the future.

It is also interesting to note that market forces have recently induced commercial software firms to adopt practices that first began in the open-source community. A new "shared source" initiative is emerging whereby governments and large corporate clients can access proprietary software code, helping to assuage fears of secret security backdoors. The code is called "shared" rather than "open", since clients are allowed to see the code but are not allowed to modify it. Shared source initiatives are expected to pick up steam in the years ahead.

In sum, open-source systems are currently too risky, and are less cost-effective than commercial software products. Proprietary software firms are also beginning to integrate the best elements of open-source solutions into their products. The Israeli government should therefore not be in a rush to abandon proprietary software. It should at least wait until the experiences of other local and national governments that have adopted open-source solutions (e.g., Munich and Peru) have been sufficiently analyzed. Israeli taxpayers are currently better off without open-source software solutions in the public sector.

Dr. Robert M. Sauer is a lecturer in the Department of Economics, Hebrew University of Jerusalem, and president of the Jerusalem Institute for Market Studies.

Published by Globes [online] - - on December 8, 2003

Twitter Facebook Linkedin RSS Newsletters גלובס Israel Business Conference 2018