Information Ethicsveryard projects > information > ethics
veryard projects > ethics > information
|ethical issues||on this page||
on other pages
|There are a number of ethical issues that can arise in
Ethics of Modelling
Privacy and Data Protection
Privacy and granularity
Discrimination and Justice
Blacklists and whitelists
Modelling Ethicsveryard projects > information management > ethics > modelling
The inclusion of particular criteria in an information model, enabling particular decision or selection processes, therefore has ethical and legal implications. If people are judged according to inaccurate or inappropriate data, and if subjective assessments are mistaken for reliable facts, the information model carries some of the responsibility for this. The analyst cannot hide behind the convenient fiction of moral neutrality.
Data Ownershipveryard projects > information management > ethics > data ownership
|Who Owns Your Mother's Maiden Name?
Privacy and Confidentiality
The ownership rights, of course, entail an identity rule - in order
to show that a record on a database belongs to me (i.e. that I have certain
rights of access and correction to the record), I have to prove that it
refers to me (and not to someone else with the same name).
Privacy and Granularityveryard projects > information management > ethics > privacy and granularity
1 A single occurrence of PERSON for each human being.
2 A single occurrence of PERSON for each human being in each (socio-economic) role.
3 Personal information aggregated into demographic or behavioural statistics.
One of the aims of the UK Data Protection Act (and of similar legislation in other countries) is to prevent the combination of data from several sources, for purposes other than that for which the data were originally collected. This means that (2) is preferred to (1). For some purposes, we are only allowed access to statistical aggregations of data, but not the raw data themselves. This means that (3) is preferred to (1) and (2).
Then the fun is to predict the behaviour of an individual from the demographic data, for example:
|Grain and Granularity|
Classification and Discriminationveryard projects > information management > ethics > discrimination
Classification of some sort is a necessary fact of life. We want to be able to discriminate between capable and incapable, safe and dangerous, polite and rude, even perhaps good and evil. But sometimes classification is arbitrary; and there may be as many classifications as there are interested parties.
The very word ‘discrimination’ is often used to denote unjust or unfavourable treatment of an identifiable group of people. But it is a fallacy to think that classification itself, or discrimination are themselves inherently undesirable. After all, unfair discrimination can only be recognized (let alone corrected) by a similar (but not necessarily equally unfair) discrimination: if a black woman engineer wishes to prove that she has been unfairly discriminated against because of her colour or sex or qualifications, she must herself classify herself in this way.
Classification of people is not just a necessary evil, but is (most
of the time) a useful and acceptable procedure. Consider schools,
for example, where at first sight the word ‘class’ appears to have a rather
different meaning. However, schoolchildren are divided into classes
by some classification, based on age (usually), sex (often), ability (possibly),
mother tongue (perhaps), or some other characteristics. (For the
rich, for the physically disabled or musically talented, and for religious
minorities, there may even be separate schools.) Too great a diversity
of children within a class makes it impossible for the teacher to communicate
effectively with the whole class. Educationalists may argue at great
length exactly which characteristics should be used, and exactly how much
diversity or uniformity is desirable, but few of them would expect a 16-year-old
bookworm, who spoke three languages but whose English was rudimentary,
to be forced to learn alongside a 9-year-old who couldn’t yet read, and
only spoke English.
|Classification as form of Abstraction|
Who controls the criteria by which you will be judged?veryard projects > information management > ethics > criterion ownership
You cannot (or at least should not) introduce new criteria of relevance. (This is why retrospective legislation is frowned upon.) For example, if an employee is told that the criteria for promotion to the next grade are such-and-such, and then told (after struggling to attain the set standard) that the criteria have now changed (this is known as moving the goalposts), s/he may be disheartened.
Collecting additional information about employees’ performance may inadvertently have this effect. Employees (and their representatives) can sometimes be very suspicious of additional performance measurement, because they expect it to influence and alter the existing criteria for promotion and remuneration.
Do you have a secret past?veryard projects > information management > ethics > secret past
Information, at least as defined by computer programmers and bureaucrats, is intolerant of ambiguity. People are defined by a relatively small number of attributes, in a fixed way. A rash act at university, or in adolescence, can dog a man for life. You are what you have once done:- that’s a very dangerous doctrine. It labels people permanently as subversives, communists, homosexuals, drug addicts or whatever. (It was only in 1870, according to Foucault, that Carl Westphal transformed the practice of sodomy to the state of homosexuality. “The sodomite had been a temporary aberration; the homosexual was now a species.”)
Under English law, certain types of information are excluded at certain stages of a criminal trial. Previous offences are not to be mentioned until the verdict has been reached, but may be considered in deciding the sentence. However, minor offences are supposed to be deleted from the record altogether after a set time period.
This is a deliberate attempt to separate the individual from his/her historical behaviour. This allows an individual to be distanced from the past. Otherwise a criminal record can act as a millstone, preventing even the truly repentant criminal from ever going straight, because nobody will give him a straight job, or a mortgage, or an overdraft, or decent housing, or whatever.
It is a difficult moral decision. If A trusts B with the care of children, or money, or whatever, should A be informed that B has once been convicted of abusing a similar trust in the past? It's easy to say yes to that question - but suppose that B has never been convicted, but has been tried and acquitted (thanks to insufficient evidence, or perhaps some legal technicality). Now suppose that B has never been tried, but has been questioned by police on several occasions, and there have been unsupported allegations in the press. Should all of this information be public knowledge? What aspects of B’s past history is A entitled to know, and what aspects is B entitled to put behind him/her? Should B be allowed to bury the past and rebuild his/her reputation?
Surely B should be allowed to forget minor breaches of the law, committed many years ago and properly punished, provided that B has not lapsed since. And surely it is unfair to repeatedly confront B with unproven and possibly malicious allegations and accusations. Howewer, there may well be some incidents that B should never be allowed to forget or conceal.
According to English libel law, it is insufficient defence against a
charge of libel that what is published happens to be true. Thus truth and
honesty must be weighed against reputation. The American approach is different.
Several candidates have been forced to drop out of the race for the Presidency,
their reputations sullied by the uncovering of sometimes quite bygone indiscretions.
Blacklists and whitelistsveryard projects > information management > ethics > blacklists
To evade such controls, some organizations now use whitelists. A whitelist includes everyone that is an acceptable credit risk or security risk. Exclusion from a whitelist has exactly the same effect as inclusion in a blacklist, except that there is no legal mechanism for redress.
Implementing a whitelist solely in order to evade these controls is immoral. Any organization using either whitelists or blacklists has a moral responsibility to maintain the accuracy of the list, and to enable any valid correction process.
If you have to have a list at all, probably the best solution from an information modelling point of view, as well as from an ethical point of view, is to combine blacklist and whitelist into a single list, based on the entity type PERSON, differentiated by an attribute indicating whether s/he is acceptable or unacceptable. Furthermore, the evidence supporting acceptability or unacceptability should be preserved, to provide an audit trail enabling efficient and effective correction. There should also be a mechanism for deleting past misdemeanours, as discussed above.
|veryard projects > information management > ethics||
Copyright © 1992, 2001 Veryard Projects Ltd