Hazard algorithms were already reported in the 80s

The advent of the personal computer

In contrast to now, there was not always broad political and social awareness of digitalisation. In the 1980s, the decade when centralized computing centers gave way to the personal computer and the automation of governments and other organizations boomed, there is little trace of discussion of computer-related topics.

“Politicians, administrators and civil servants liked to talk about all sorts of topics, but the topic of digitization, or, as it was called then, automation, was rarely discussed. This despite the fact that public investments in IT had already started in the middle of the 1980s amounted to more than 2 billion guilders a year, and there had been a state secretary with automation in his post much earlier, in 1967,” says Ter Bekke. “Wim van Gelder, who was deputy in North Holland in 1984 and took charge of the information and automation portfolio, noted that he only got this topic because none of his fellow directors wanted it. And Hans Wiegel was undoubtedly right too. when he said that during his time as a minister in the Van Agt cabinet there was “no political gain from automation.”

‘The greatest benefit is the greatest danger’

Despite the fact that there is little interest in the subject of digitization from a political-administrative perspective, Ter Bekke found articles in Domestic Governance where digitization is discussed in relation to themes such as the rule of law and form of government. For example, editor Tom Eijsbouts reports in the issue of 1. February 1985 about a symposium ‘Trias Automatica’ at Tilburg University. The symposium is chaired by Herman Tjeenk-Willink, then government commissioner for the reorganization of the civil service. The meeting is about the relationship between automation and the rule of law, and the presenters approach the topic not from technology, but from law and governance.

The central question at the symposium is what the consequences will be if the government starts making decisions, or more precisely, makes decisions based on digitally collected data, and uses algorithms to make those decisions. Gert Nielen, professor of information systems, is in any case not completely reassured. “You no longer look at that data, and those things take on a life of their own. What is recorded in a computer program no longer requires our daily attention. That is the greatest advantage of automation, but at the same time its greatest danger.”


Another point raised by the speakers has to do with the danger of the government acting unilaterally. Ernst Hirsch-Ballin, professor of constitutional and administrative law at the time, notes that too much reliance on one’s own, automatically collected data when making decisions can lead to isolation from bureaucracy and the blurring of the legal principle of ‘fair hearing’ ‘.

Additionally, he argues, the ability to forget, or prescription, is a basis for law, but not a property of computers. Nijmegen professor of public administration Hans Bekke is also afraid of a one-sided government. According to him, automation may lead to the government preferring to build ‘silver-plated’ data files rather than collecting its information from intermediary organisations. All you’re left with, he says, “is government services each bombarding each other with their own figures”.


A third concern has to do with the possibility that algorithms will be used to make decisions. “Legal reasoning,” Hirsch-Ballin argues, “is a process that eludes the algorithm. It embraces unpredictability, unpredictability that is not randomness.” Where there is a need for consideration, where there is a power and no government duty to make decisions, there are limits to the use of computers, he believes.

“Such discretionary powers are intended to provide justice to the individual case, and in such cases judges also ask for a judgment directed at the individual situation. Algorithms are therefore of no use here,” writes Ter Bekke.

Additional matter?

Although it is a stretch to attribute prophetic gifts to the speakers of the past, it is remarkable how relevant their analyzes have proven to be for the present. Everything that went wrong in the welfare affair was pretty well foretold at that meeting in 1985. Algorithms generating decisions, data files taking on a life of their own, the inability to do justice to individual cases, and judges not offering any corresponding power. That it was already possible to make such an accurate analysis of the role of digitization in the rule of law in 1985, when PCs were still a sight and the Internet revolution had yet to begin, is quite an achievement.

This article was previously published by Domestic Governance.

Read the full essay (Digital Lessons from the Last Century) in this week’s Domestic Governance No. 18.

Leave a Comment