The Spam Bill 2003 – Response to EFA Criticism
Introduction
Electronic Frontiers Australia recently came out with a detailed criticism of the Spam Bill 2003, in which they describe it as “not anti-spam”. After analysing the EFA criticisms, CAUBE has concluded that this label is entirely unfair. Most of the criticisms made regarding the effect of the Bill simply do not bear out. Furthermore, the criticisms do not go to the core of the Bill, but rather to the fringes.
To begin, we need to look at the structure of the Bill. There are 47 clauses in the body of the Bill. There are also three schedules, with 5, 6 and 10 clauses respectively. There are also powers to make regulations in many parts of the Bill.
EFA's criticisms of the Bill all relate to the content of the schedules. The schedules in this Bill are somewhat unusual as they contain substantive rules that you would normally find in the body of the Bill. The reason for putting substantive provisions in a schedule is normally to be able to replace those rules more easily later by simply replacing the schedule. Placing these items in a schedule is done because you acknowledge that if there are problems, that's where they are. What's more, each of the schedules includes a regulation-making power to allow the Government to respond quickly if problems do arise. Accordingly it is somewhat misleading to look at rules in the schedules alone and then claim an undesired outcome will eventuate.
Many of the concerns are matters capable of being dealt with by making regulations.
Many of the points in the EFA analysis are repetitive. For example, where a point is made about an exemption generally, it is also made in the context of each provision to which the exemption applies. Rather than make redundant responses, we have endeavoured to respond to the key points made.
The Manner of Banning Spam
The ban on spam in the Bill is achieved by first banning unsolicited commercial email (UCE), and then making exemptions in circumstances where it would be prohibitively difficult to use UCE in bulk. There are also exemptions that are, broadly speaking “policy” exemptions, which are made in situations where spam is unlikely to be used in any event. The effect of the Bill will be to ban unsolicited bulk commercial email (UBCE), with very little non-bulk commercial email banned.
The Bill has also been designed as a proposed template for international adoption. The sub-category of “UBCE” is more amenable to this goal than either the “UBE” or “UCE” categories alone. While there is spam that is not commercial, the vast bulk of spam is in fact commercial, and this Bill clearly sets the appropriate – opt-in – standard for such spam.
It may well be that once UBCE is banned around the world, non-commercial UBE will increase in volume. If and when this occurs, the opt-in standard in the Spam Bill 2003 will be the starting point for further legislative changes in Australia.
There is no doubt that the effect of the Bill is to ban spam. Not all spam, but the vast majority of spam. It is designed to, and will in fact, ban the largest categories of spam. It will shut down Australia's two most prolific spammers, as well as several smaller scale operations scattered around the country. This is not a case like dominant Bills in the United States, which are opt-out. It is fair to describe opt-out bills as “not anti-spam”, but the Spam Bill 2003 is opt-in. It sets the right base standard to be used going forward, and the peripheral matters impugned by the EFA are matters to be adjusted by regulation and by adjustments to the Act down the track.
The Review
The Bill contains a provision (clause 46) requiring the Minister to conduct a review of the operation of the Act before the end of two years and four months after Assent. That review is likely to focus in a large part on the effect of the schedules. This is clearly not intended to be a “set and forget” solution. The Government has made every effort to ensure that the Act will be adjusted as necessary to reflect changing or unexpected circumstances.
The Case Scenarios
The first set of EFA criticisms centre around examples or “case scenarios” in which the Bill allegedly produces absurd results. Accordingly, we will analyse the examples and explain in each case where the example fails.
Case Scenario 1 (under Issues 1)
An individual has a personal (not business) web site and publishes their resume on it with their personal (not work-related) email address, for example: xybloggs@yahoo.com.au. The individual may, for example, be an unemployed person, or an employed person who is nevertheless interested in full or part-time contract, consultancy or job offers.
Another person (or organisation/company) wishes to email the individual to offer a business opportunity, for example, a contract for work that is directly relevant to the experience and skills set out in the individual's resume.
The message would be caught by the definition of "commercial electronic message" because its purpose is to offer a business opportunity. A contract for work/consultancy or employment would constitute a "business opportunity" - if it would not, then neither would many messages that unquestionably are spam such as: "Work from home selling this or that"; "Make money advertising these porn sites on your own site"; etc, etc.
...
"Inferred consent" means consent that may reasonably be inferred from the conduct and the business and other (e.g. family) relationships of the individual or organisation concerned. Hence, unless there is a prior relationship between the sender (person or organisation) and recipient, consent as currently defined in the Bill cannot be inferred.
This example has two problems. Firstly, it misunderstands the legal meaning of “business opportunity”. Secondly, it misunderstands the nature and effect of the conjunction in clause 2(b) of schedule 2 relating to inferred consent.
“Business” is usually defined by reference to the “indicia of business”. They generally require some sort of commercial endeavour (Commissioner of Taxation v Bivona (1989) 89 ATC 4183). The indicia include a profit-making purpose, repetition and regularity, organisation and system, size and scale of operations, and other factors of a business nature (Ferguson v FCT (1979) 9 ATR 873; 79 ATC 4261). Other factors may include the type and quantity of goods traded (Edwards v Bairstow [1956] AC 14) and the use of a company (Lewis Emanuel & Son Ltd v White (H M Inspector of Taxes) (1965) 42 TC 369; London Australia Investment Co v FCT (1977) 7 ATR 757). Regardless, employment is not normally considered a business opportunity, and even if done by means of a contract is unlikely to entail a business unless it exhibits some of the indicia – most notably repetition, regularity, organisation and system. On the other hand, if sufficient indicia are present, then the business function exemption will most likely apply.
In the case of “work from home” offers, several indicia are likely to be present: profit-making purpose; repetition and regularity; trading; organisation and system.
Regarding the definition of “inferred consent”, while it is true that the consent must be reasonably inferred from “(i) the conduct; and (ii) the business and other relationships” of the individual or organisation concerned, EFA's understanding of the conjunction is flawed. While an “and” conjunction in legal analysis normally means you examine each side of the conjunction independently and require both to be satisfied, where the conjunction is used to join factors to be considered, it means you consider all of the factors applicable to the case, taking into account the effect that each factor has on the others. See, for example, the Contracts Review Act 1980 (NSW) s9, as interpreted in West v AGC (Advances) Ltd (1986) 5 NSWLR 610. In this case the fact that the résumé has been published the web site is such a powerful invitation to make a job offer that it is unlikely any part of the “business and other relationships” factor could override it.
This case is unlikely to be within the term “business opportunity”, but quite clearly falls within the “inferred consent” exemption. The example is not even arguable.
Case Scenario 2 (under Issues 1)
An individual has a personal web site containing a number of articles written by them about one or more topics.
A publisher of a magazine or author of a proposed book wishes to email the individual offering to pay them for the right to re-publish one of their articles in a magazine or book.
Such a message would be caught by the proposed law because the purpose of the message is to offer a business opportunity, and the same situation applies as in Case Scenario 1 above, irrespective that the particular individual would (most probably) want to receive that message and offer.
Leaving aside the fact that if “the particular individual would (most probably) want to receive that message and offer”, this is an indication of consent being inferred from conduct, there are clearly no indicia of business here. In law, this is what would be called a one-off sale of a capital asset. There is no profit making plan on the part of the author, no repetition or regularity, no system or organisation, no large scale operation, no trading, no corporation. That is, not one of the indicia normally used is present.
This case example is not even arguable.
Case Scenario 3 (under Issues 1)
An individual has a personal (not business) web site providing information about a particular topic and also publishes a list of recommended books on the same topic. The individual's personal (not work related) email address, for example: xybloggs@yahoo.com.au, is also available on the site.
A new book about exactly the same topic is published and the author's public relations company wishes to send the individual an email message offering the person a copy of the new book.
Such a message would be caught by the proposed law because the purpose of the message is clearly to promote a book (that is, a good as in "goods and services") and the same situation applies as in Case Scenario 1 above, irrespective that the particular web site owner would (most probably) want to receive that message and offer.
Once again, the conduct of the individual would bring this within the “inferred consent” exemption. Indeed the reason the example seems so objectionable if accurate is that common sense suggests it is within the inferred consent. They have published their email address on the web site associated with a particular topic and consequently must be expecting relevant messages related to that topic or to the web site.
This is probably the most arguable of the three examples, but it is still comfortably within the concept of inferred consent.
Summary of the Case Scenarios
Even though none of these cases holds up to close scrutiny, the Government has planned well for the possibility of unforeseen effects in this area. There is a regulation power in clause 6(7) of the Bill which allows certain types of messages to be excluded from the definition of “commercial electronic messages”. Thus anything inadvertently caught can be quickly removed by regulation. Additionally in clause 5 of schedule 2 there is a regulation power to make rules defining certain things to be inferred consent, and certain things not to be inferred consent. As if that were not enough, the rules on consent themselves are in a schedule, which allows for them to be easily replaced if the present rules are found not to be workable.
The Exemptions
The exemptions are, broadly speaking, policy based exemptions. While CAUBE does not necessarily agree with all the exemptions, we note that they are placed in a schedule so they can be replaced easily in response to later developments.
Exempt Message Class – Factual Information
The intent and effect of this exemption (schedule 1, clause 2), is that a message should not be classed as a commercial electronic message solely by reason of standard information in the message's headers and the signature portion. It operates by making those parts of the message irrelevant for determining if the message is a commercial electronic message. If the “factual content” is information about a commercial product or service, the message will still be commercial, even though the content is “factual”. This is an important exemption when it is “UCE” that is being banned, otherwise almost every email message currently sent from a business would be banned.
It may well be that this exemption will need some adjustment, but it would be entirely wrong to remove it. The government has clearly acknowledged the potential need for adjustment by including a specific regulation making power in clause 2(c) of schedule 1 that allows additional conditions to be placed on messages in this category. For instance, one additional requirement may be that the content other than the exempted content be such that it is reasonable to conclude that the message would have been sent even if it did not contain the exempted content.
Exempt Senders – Government, Political Parties, Religious Organisations and Charities
While these organisations are exempted (schedule 1, clause 3), the impact of the exemption is likely to be minimal. CAUBE believes some or all of these exemptions will need to be removed at a later date, but that their inclusion at this stage is not fatal to the Bill. Political parties, governments (at least at the Ministry level) and charities in particular are unlikely to use UBE because they rely on broad support and cannot afford to upset the volume of people who are likely to react adversely to UBE. While UBE has been used by some political candidates in the United States, there is no evidence that this has helped their campaigns, and it is far more likely that it damages their campaigns.
In all cases, these are groups that rarely offer products or services of a nature that lends itself to spam. They in fact rarely spam. If and when this situation changes, the schedule can be replaced to remove the exemption.
Exempt Senders with Limited Target Audiences – Educational Institutions
This exemption (schedule 1, clause 4) is limited in scope and has the effect of exempting messages sent to the household of a person who studies at that institution. This exemption is useless for most spam, and the cases in which it might be used would normally be covered by the Privacy Act 1988 (Cth). While CAUBE would not necessarily agree with this exemption, its impact is minimal and not something that should prevent the Bill from moving forward.
Consent by Conspicuous Publication
EFA claims that the inferral of consent by conspicuous publication allows small business owners to be spammed with a vast range of things. In fact the spammer would have to visit the web site, find the email address and determine if it met the conception of conspicuous publication, and check to see that no message indicating a desire not to receive spam is present. Thus this form of consent is useless for bulk.
Remedies
EFA claims that the penalties that can be imposed by means of the infringement notice are inappropriately high for a first offence. This fails to have regard to several facts and some issues of law. The Bill also grants a power to issue formal warnings or obtain consent orders without issuing an infringement notice. Given this, there is a discretion to be exercised. If the ACA were to issue an infringement notice for every first and trivial offence, they would likely not have the funds to pursue the matter in Court. Furthermore, the decision to issue such a notice would be capable of being challenged on administrative grounds – notably, failure to take into account relevant considerations, taking into account irrelevant considerations, or Wednesbury unreasonableness.
The Consequential Amendments Bill
While the provisions of the consequential amendments Bill are outside of the areas that CAUBE normally deals with, we would note that the powers in those provisions are not unusual in Commonwealth Regulation. Compared to the powers of some investigation and enforcement agencies, including the Australian Taxation Office and the Australian Competition and Consumer Commission, they are positively mild.
The strongest complaint of EFA is that the provisions on consent to searches would allow either a flatmate or a landlord to consent to the search. A flatmate can consent to such a search at common law, although this might not extend to searching the computer depending on all the circumstances. A landlord cannot consent at common law, although an owner of unoccupied premises can.
The provisions allowing for a search by consent are primarily necessary because of the limitations on Commonwealth power. The provisions have the effect of allowing the ACA to act on consent given, and we would not take them to empower the flatmate or the landlord to give a consent that they could not give at common law.
The powers regarding search warrants would also be subject to the common law rules regulating the issuing of search warrants, which include requirements that the issue and scope of the warrant is necessary and reasonable. Most notably, such a warrant could not issue if the evidence could be obtained without the warrant. Even in the unlikely event that the ACA thought it fit to serve a warrant on a recipient of a spam (unlikely because they would first have to identify a particular recipient and have a need for the recipient's copy as evidence) they would still need reason to believe that the recipient would refuse to supply the evidence.
Conclusion
The Spam Bill 2003 sets the right base standard – opt-in. There are some types of spam that are excluded for the time being, but this can be adjusted as necessary and in line with that base standard. The specific examples given by EFA of non-spam being caught do not hold up to scrutiny, and although it may be possible to identify non-spam that would be caught, there are extensive regulation-making powers that allow for this situation to be dealt with quickly.
This is not a situation in which disputes at the periphery of the Bill should prevent the Bill going forward. While there is no doubt room for improvement, the facilities to make such improvements are included in the Bill. While there are some areas of policy, notably in the exemptions, where CAUBE would like to see adjustments, it is abundantly clear that the Bill as drafted is suitable for enactment.