Archive for the ‘Government’ Category

Browne Report and graduate tax

Oct 14, 2010 1 comment

The Browne report, Securing a Sustainable Future for Higher Education, has just been published.

The almost universal view among commentators is that Lord Browne has conclusively argued the case against a graduate tax.

In this blogpost I argue that the Browne Report has done no such thing. Rather it has set up a straw man version of a graduate tax, and shown this straw man to be unworkable. It is perfectly possible to set up a version of a graduate tax that avoids the Browne Report’s objections. (Note that in this post I am not arguing for or against a graduate tax, merely that the Browne Report has not made a convincing argument against a graduate tax.)

[A straw man argument is an informal fallacy based on misrepresentation of an opponent's position. To "attack a straw man" is to create the illusion of having refuted a proposition by substituting a superficially similar yet unequivalent proposition (the "straw man"), and refuting it, without ever having actually refuted the original position.]

What some of the commentators say

The Times in its editorial of Wednesday 13 October 2010 says:

This series of proposals demolishes the case for centralised control of universities and they demolish the case for a graduate tax…

Vince Cable in his statement to the House of Commons says:

But Browne identifies serious problems with a ‘pure’ graduate tax. The proposal is unworkable; does not produce sufficient revenue to finance higher education until 30 years from now; weakens university independence; and is unfair to British graduates as opposed to graduates living overseas.

The Independent in its editorial of Monday the 11th October says:

Vince Cable has, mercifully, surrendered the notion of a graduate tax, paid retrospectively, which would have been a bureaucratic nightmare to administer.

The Guardian in its editorial of Wednesday 13 October 2010 says:

Browne’s plan, with its strong support for the less well-off, is also better than a full-on graduate tax.

Have any of these people actually read the Browne Report?

What the Browne Report actually says

The arguments against a graduate tax are in Chapter 7, Other Proposals we have Considered. I’ve repeated them below for reference:

Issues Graduate Tax Student Finance Plan
1 Cost of learning No upfront costs No upfront costs
2 Cost of living Will require support through loans – this means that graduates have to pay the additional tax as well as make loan payments Graduates make a single set of payments to cover the costs of learning and living provided upfront by Government
3 Payment terms Linked to income, no fixed mortgage style payments, payments continue indefinitely Linked to income, no fixed mortgage-style payments , payments stop when costs of learning and living are paid back – or 30 years – whichever is earlier
4 Protection for graduates on low incomes Graduates start paying when they cross the income tax threshold – £6,475 per year Graduates pay nothing until they earn £21,000 per year
5 Costs for graduates Uncapped, could be several multiples of the cost of the degree Maximum payment is equal to the charge for the degree. Majority of graduates will pay less
6 Funding to universities Tax revenues take time to build up – for first 25 years, model depends on Government filling that gap; after that, depends on Government enforcing a ring fence around graduate tax revenues Direct funding relationship between student and university
7 Burden on Government Additional £3bn a year until 2015-16 at least; additional spending continues until ca. 2041-42 No additional spending; continuing requirement to provide student finance
8 Relationship between students and universities Student experience does not matter to university for raising funding University depends on student willingness to pay for significant proportion of funding, so providing a high quality student experience is critical
9 Incentives for institution No variability in funding, so no incentives to focus on quality, access or student experience Sustaining income – or raising it – depends on improving quality, access and student experience

Virtually all of these objections assume a particular (and not very sensible) version of a graduate tax.

Lets look at each in turn:

1) Cost of learning – same for both options.
2) Cost of living – true, but the level of the graduate tax could be set to take this difference into account.
3) Payment terms – there is nothing inherent in a graduate tax that says it must continue indefinitely. It be formulated to stop when the tuition costs have been met and/or after 30 years. (Indeed, the Browne Report even mentioned it looked at the NUS proposal of a graduate tax with a capped maximum contribution.)
4) Protection for graduates on low incomes. Again there is nothing inherent in a graduate tax that says it must kick in as soon as the graduate starts to pay income tax – it could also kick in at an earnings threshold of £21,000.
6) Funding to universities – clearly a graduate tax cannot immediately replace the current system, it needs to be phased in over a number of years.
7) Burden on Government. As for (6) a transition period is clearly required. The Browne Report does not make it clear why the burden to government during that transition period would be any more than the current system.
8) Relationship between students and universities. The Brown Report’s statements are true if the graduate tax is paid to central government. However the report has not even considered the possibility of the graduate tax being paid directly to the university. Such a tie would mean the university would have a long term interest in the success of the student.
9) Incentives for institution. Similarly to (8), if the graduate tax is paid directly to the university, then the university has a long term interest in the student. This would provide incentives not only for good teaching, but also good careers advice and even for continuing education and refresher courses. It would also provide disincentives for admitting students just to make up the numbers, since if these students failed to graduate, the university would not gain any income from them.

Of all the arguments posed against a graduate tax, the only serious one is about how to manage and fund the transition from the current system to a graduate tax. The Browne Report doesn’t consider a transition period – it looks at an immediate changeover to a graduate tax with the funding gap filled by government: of course this is going to look unfavourable, it’s the most expensive (to government) way of funding the transition.


The Browne Report’s case against a graduate tax is based on a fallacious straw-man argument.

There may be legitimate arguments against a graduate tax, but the Browne Report has not made them.

The First Amendment and religious freedom

Aug 11, 2010 1 comment

Here is the First Amendment in full:

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

Commentators tend to state that religious freedom is protected by the establishment and free exercise clauses of the First Amendment. For example the website 1 for All, says in its FAQ:

5. How does the First Amendment protect religious liberty?
It protects religious liberty through the establishment clause and the free exercise clause. The establishment clause — “Congress shall make no law respecting an establishment of religion” — provides for separation between church and state. The free exercise clause — “or the free exercise thereof” — means that individuals can hold whatever beliefs they wish on religion or nonreligion and to freely practice those beliefs.

While this is certainly true, it’s not the whole truth. These two clauses certainly explicitly protect religious freedom; all the other clauses, to some extent, implicitly protect religious freedom.

To further discuss this, I’ve parsed the First Amendment into its separate clauses, which I discuss separately. (Note that parsing the First Amendment is a non-trivial exercise and there is at least one other parsing that I considered – if you disagree with my parsing below, I’m most interested in hearing your version.)

Congress shall make no law respecting an establishment of religion

One of the two clauses explicitly protecting religious freedom.

Congress shall make no law prohibiting the free exercise of religion

One of the two clauses explicitly protecting religious freedom.

Congress shall make no law abridging the freedom of speech

Nowadays, when we talk about freedom of speech, we often think of freedom of political speech, but free speech is required for the practice of most religions. Free speech protects the right to preach, to pray aloud, to recite creeds and simply to state one’s religious beliefs.

Congress shall make no law abridging the freedom of the press

Nowadays, when we talk about the press there is almost an implicit assumption that we are talking about the Fourth Estate, that is newspapers and other journals. The framers of the Bill of Rights certainly saw a free press as a protection against tyranny. For example, George Mason in his Master Draft of the Bill of Rights stated “the Freedom of the Press is one of the great Bulwarks of Liberty”.

However the term “press” has a wider meaning, it encompasses the printing press and also applies to any form of publishing establishment.

A free press is required for the printing of religious texts. Restriction of the press has been used to restrict the freedom of religion, and the framers of the Bill of Rights would certainly have been aware of this. For example:

In 16th century England the Tyndale Bible, the first mass produced English translation of the Bible, was banned. Tyndale was arrested by church authorities, tried for heresy, strangled and burnt at the stake.

In 1668, after writing The Sandy Foundation Shaken (a text espousing his Quaker views) William Penn was charged for publication without a license and imprisoned in the Tower of London.

Congress shall make no law abridging the right of the people peaceably to assemble

Nowadays, when we think about the right to assemble, we normally think of political assembly and in particular the right to protest. But we shouldn’t forget that the practice of most religions involves peaceful assembly. Religious persecution has often been carried out by breaking up religious meetings, prosecuting those involved in said meeting, or banning such meetings. For example:

In 1662 John Browne was arrested was arrested on orders of Peter Stuyvesant, Director-General of the colony of New Netherland (now New York) for allowing a Quaker meeting in his house.

In England the Conventicle Act of 1664 forbade religious assemblies of more than five people outside the Church of England.

In 1670 William Penn was arrested, accused of preaching before a gathering in the street. Penn had deliberately provoked the authorities in this way to test the validity of the then new law against assembly.

Congress shall make no law abridging the right to petition the Government for a redress of grievances

What’s this about then? It seems almost unrelated to the other clauses.

In December 1657 by a group of citizens from the town of Flushing (now part of Queens, New York) petitioned Peter Stuyvesant, protesting against his persecution of Quakers. None of the 30 signatories of the petition were Quakers themselves. This petition became known as the
Flushing Remonstrance. Stuyvesant arrested those who presented the document to him, and forced the signatories to recant. Edward Hart, the town clerk, and Tobias Feake, sheriff of Flushing, refused to recant, and spent over a month in prison.

Nowadays we tend to think about the right to petition as a means of achieving political aims and as a means of protecting public participation in government, however petitions have historically been used to protest against religious persecution.

In summary

Every single clause of the First Amendment protects religious liberty. Removal of any of these clauses would open the door to a form of religious persecution.

Young Rewired State and government electricity wastage

Aug 9, 2010 Leave a comment

I was recently peripherally involved in Young Rewired State 2010. This was a week long event for young developers aged 15 to 18 and was held across the UK. The idea was simple: the young developers had five days to create an application or website based on a theme that interested them – the only proviso was that it had to use at least one government dataset.

In the Osmosoft offices we hosted a group of seven young developers: Issabell, Daniel, Harry, David, Dylan, Marcus and Sufian. Between them they produced a variety of applications:

Live London Leisure Locator, by Dylan Maryk
CycleHubs, by Daniel May
Un-Named, by Sufian Hassan and Dylan Maryk
GovSpark, by Isabell Long
Psych Survival, Rates by David Goater
UniSearch, by Sufian, David Goater, Harry Burt, Daniel May and Marcus Hughes.

The full list of projects is here.

Of these projects, the one I found most interesting was GovSpark by Isabell Long. A number of government buildings now provide downloadable files of their historical energy use. For example, the Department of Energy and Climate Change’s data is here. Isabell used this data to compare the daytime and nighttime electricity usage of various government buildings. Her investigation showed that many government buildings use substantial amounts of electricity at night – a nighttime electricity consumption rate of 40% of the daytime rate is not atypical. This potentially represents a huge waste – a waste of electricity and also a waste of taxpayers money. Although there may be legitimate reasons why government buildings use so much electricity overnight, I suspect the main reason is that lighting and computers are unnecessarily left on overnight.

I was happy to see that Isabell was awarded the Most likely to antagonise the CIO Council Rewired State Award.

Online Copyright Infringement and DEA consultation – my response

Jul 31, 2010 Leave a comment

Ofcom sought views on a code of practice called “the Online Copyright Infringement Initial Obligations Code” as outlined in its document Online Infringement of Copyright and the Digital Economy Act 2010. The consultation page is here. The consultation document is here.

My response:

Question 3.5: Do you agree with Ofcom’s approach to the application of the 2003 Act to ISPs outside the initial definition of Qualifying ISP? If you favour an alternative approach, can you provide detail and supporting evidence for that approach?

I do not think Ofcom’s position is sufficiently clear to either agree or disagree. Paragraph 3.23 states:
“Where a Wi-Fi network is provided in conjunction with other goods or services to a customer, such as a coffee shop or a hotel, our presumption is that the provider is within the definition of internet service provider.”

However it is not clear what is meant by provider. So for example, consider a Starbuck’s coffee shop which also provides Wi-Fi. Is the provider the individual coffee shop, or is it the entire Starbuck’s chain? Similar questions apply to hotels, libraries, airports and other providers of Wi-Fi access. Even with this issue clarified, it is not at all clear how such providers of Wi-Fi would be expected to count their subscribers, or even how a subscriber could be defined.

Question 3.6: Do you agree with Ofcom’s approach to the application of the Act to subscribers and communications providers? If you favour alternative approaches, can you provide detail and supporting evidence for those approaches?

Paragraph 3.31 states:

“Those who wish to continue to enable others to access their service will need to consider whether take steps to protect their networks against use for infringement, to avoid the consequences that may follow.”

It is entirely unreasonable to expect users to police the use of their networks. Users should not be held to account for the activities of those who use their networks without permission, whether or not that network is protected. If someone steals a tool from my garden shed and uses it to commit a crime I am not liable for that crime, whether or not my garden shed was locked.

Question 4.2: Do you agree with our proposal to use a quality assurance approach to address the accuracy and robustness of evidence gathering? If you believe that an alternative approach would be more appropriate please explain, providing supporting evidence.

The DEA states:

“124E Contents of the Initial Obligations Code
2.26 This sets out what the code underpinning the initial obligations (whether an industry code or Ofcom’s own code) must contain. The code must set out the process by which the initial obligations will operate and the procedures that copyright owners and ISPs must follow in relation to them. It must set out the criteria, evidence and standards of evidence required in a CIR and the required format and content of a notification letter sent to a subscriber. It must not permit any CIR more than 12 months old to be taken into account for the purposes of a notification.”

Note it states that the code “must set out the criteria, evidence and standards of evidence required in a CIR”. The DEA is quite clear here.

Ofcom is not discharging its responsibility to set out the criteria, evidence and standards of evidence.

Ofcom in not even saying that it will agree an industry standard code with ISPs.

Ofcom is saying that copyright owners should provide a Quality Assurance Report detailing the “steps taken to ensure the integrity and accuracy of evidence”.

Ofcom is saying that (paragraph 4.6) “it requires Copyright Owners to develop appropriate technical standards”. This is incredulous: technical standards do not lay within the domain expertise of Copyright Owners. It is much more appropriate for ISPs and Ofcom to develop the technical standards.

Ofcom also says (paragraph 4.4) “we are proposing that a Copyright Owner (or an agent acting on their behalf) should, before submitting their first CIR (and from then on an annual basis), provide Ofcom with a Quality Assurance report”. So not only can the Copyright Owner set their own standards, they don’t have to make these standards available for scrutiny until they submit their first CIR.

Ofcom says (paragraph 4.6) “it is similar to approaches that have been effective in other areas of Ofcom’s remit, e.g. ensuring the accuracy of communications providers’ metering and billing; and interference standards for TV transmission;”. This again is beyond belief, Ofcom is saying that the standard of evidence required for prosecution under the DEA should be similar to the standards required to avoid TV transmission interference. This position is also inconsistent with Ofcom’s position on CIRs. In paragraph 4.3 (in relation to CIRs) Ofcom states: “We believe that this matches the standard of evidence required by the courts in relation to civil proceedings by Copyright Owners for copyright infringement.”

So Ofcom is saying CIRs must meet “standard of evidence required by the courts in relation to civil proceedings”, yet the technical standards only need to meet the level of accuracy required for metering, billing and TV interference.

Does Ofcom not realise that a chain is only as strong as its weakest link? The overall standard of evidence in a chain is only as strong as the weakest piece of evidence in that chain.

Ofcom gives the impression that it regards the setting the criteria for evidence and standards of evidence as purely a technical matter, rather than both a technical and a legal matter. Has Ofcom even taken any legal advice on this?

Question 5.2: Do you agree with our proposal to use a quality assurance approach to address the accuracy and robustness of subscriber identification? If not, please give reasons. If you believe that an alternative approach would be more appropriate please explain, providing supporting evidence.

No. My objections are similar to my objections stated in my answer to question 4.2. In short Ofcom is not meeting its obligations under the DEA. Ofcom must “set out the criteria, evidence and standards of evidence”, it cannot defer this obligation to the ISP.

Question 7.1: Do you agree with Ofcom’s approach to subscriber appeals in the Code? If not, please provide reasons. If you would like to propose an alternative approach, please provide supporting evidence on the benefits of that approach.

Paragraph 7.8 states: “In exceptional circumstances it may be appropriate for oral hearings to occur and the appeals body will therefore be able to order a hearing.”

The subscriber should be able to request a hearing. If this request is denied the appeals body will have to justify why the hearing was denied.

In order to ascertain it there might be some systematic error in the process, subscribers must be able to request a list of all CIRs submitted by all Copyright Owners in relation to that subscriber. This would allow a subscribers to ascertain, for example, if some of the alleged infringements occurred when the subscriber was not at home.

Suggestion for a cheaper Trident

Jul 17, 2010 1 comment

I share the opinion that Field Marshal Lord Bramall, General Lord Ramsbotham, and General Sir Hugh Beach expressed in a letter to the Times: that Britain’s defence needs would be best served by scrapping Trident and using the monies saved to improve Britain’s conventional forces. However, I am realistic enough to accept that this is unlikely to happen in the current parliament, especially given the statement: “The Government is committed to retaining Trident and the programme will be scrutinised for value for money” (see Strategic Defence and Security Review).

First some background on Trident: the Trident Programme is the UK’s nuclear defence programme. It consists of four Vanguard class submarines each of which carries up to 16 Trident D-5 nuclear missiles. Each missile carries 3 warheads and has a range of approximately 7,500 miles (12,000km). The Trident Programme has a 30-year lifespan that is due to end in 2024 (source: BBC Trident missile factfile). The life span of the Trident D-5 missile has been extended to 2030 when it is due to be replaced by the Trident E-6 missile (source: Missilethreat).

The submarine based system was part of the cold war strategy of Mutual Assured Destruction: no nation will lauch a pre-emptive first strike against the UK because the UK will be able to launch a devastating retaliatory strike from its submarines (since at least one of the submarines is always at sea at a hidden location).

The cold war is over. The nuclear threat against the UK is no longer that of a pre-emptive first strike, so a submarine based nuclear deterrent is no longer required.

This suggests a cheaper alternative to the full Trident Programme: move the Trident missiles to land-based silos and scrap the submarines.

There are probably technical difficulties with launching a Trident missile from a land-based silo: the feasibility and cost of a land-based Trident nuclear deterrent needs to be evaluated and compared with the option of replacing the four Vanguard submarines. Both the replacement and subsequent running costs need to be compared. The question of how many land-based missiles are required, and how many warheads each missile carries needs to be addressed (I believe that under arms limitation treaties more warheads are allowed on ICBMs than SLBMs). Currently the UK has a guaranteed launch capability (provided by the one submarine that is always on patrol) of 48 warheads provided by the 16 missiles aboard a Vanguard submarine. A similar capacity could be provided by 12 land-based missiles each carrying 4 warheads, or 10 land-based missiles each carrying 5 warheads, if such configurations are allowed by international treaties.

Moving the Trident missiles to land-based silos provides an effective nuclear deterrent until 2030. At that time there will be an option to replace the Trident D-5 missiles with Trident E-6 missiles. A land-based Trident programme is likely to be more cost-effective than a replacement submarine based programme and deserves serious attention.

I have submitted this suggestion to the HM Treasury Spending Challenge site.


A middle way on Trident, by Shirley Williams.

Airport body scanners consultation – my response

Jun 26, 2010 8 comments

The Department for Transport has invited responses from stakeholders on the Government’s interim code of practice for the acceptable use of advanced imaging technology (body scanners) in an aviation security environment. The consultation page is here. The consultation document is here.

My response:

This document, dated 18th June 2010, contains my response to Department for Transport’s consultation “Code of practice for the acceptable use of advanced imaging technology (body scanners) in an aviation security environment” at The consultation paper itself is at

I am commenting as an individual. My name and address is:
[address removed]
Please remove my address from any published form of this document.

The questions posed in the consultation document are fairly narrow in scope, and miss out some important issues related to the deployment of scanners. For completeness I have answered the questions, however it is important to note that the most important part of my response is in the paragraphs before my answers to the questions.

The decision to introduce and roll out body scanners at UK airports was a knee-jerk reaction in response to a single aviation incident. The trouble with such reactions is that they are made quickly and so there is no time to do a proper analysis and give a considered response. The maxim “act in haste, repent at leisure” applies. We’ve just had one of these ill-thought out schemes put on hold: the home secretary, Theresa May, has just scaled back the child worker vetting scheme – a scheme that was hastily concocted after the Soham murders in 2002. Rather than hastily add yet another ad-hoc airport security measure, we should take the time to make a thorough analysis of the threats to air travel and update our airport security measures accordingly. As security expert Bruce Schneier says “there’s been far too little discussion about what worked and what didn’t, and what will and will not make us safer in the future”.

The introduction of airport scanners is just the latest step in a sequence of piecemeal, reactive, retrospective security measures. There was a shoe bomber, so we start to examine people’s shoes; there was a threat of a liquid bomb, so we ban liquids; there was an underpants bomber, so we decide to introduce technology that enables us to examine people’s underwear. Such an approach is doomed to failure, since it will fail to anticipate the next bomber. Schneier calls this “magical thinking”: “If we defend against what the terrorists did last time, we’ll somehow defend against what they do next time. Of course this doesn’t work”. I repeat: what we need is a proper holistic review of airport security. Let’s take time to make a considered, thorough review of the threats and issues involved, and then act on the recommendations of that review.

All security measures are a trade-off. Whether the trade-off is worthwhile depends on a number of factors including the costs (including non-monetary costs such as inconvenience) and efficacy of the security measures, the risks of the security measures failing, and the consequences of a security failure. The consultation document looks at some of the non-monetary costs of airport scanners (for example, privacy invasion, health risks and risks of discrimination), but fails to consider either the efficacy of scanners or their monetary costs. The following two paragraphs look at these issues.

Efficacy. There is anecdotal evidence that airport scanners “don’t do what it says on the tin”. On a German television program a physics professor demonstrated how to get all the ingredients of an explosive device through a scanner, see: . The “Independent on Sunday” asserts that airport scanners would not have detected the underwear bomber, see “ . If scanners don’t actually work reliably, then all the health and safety, privacy, and cost issues are moot – they are simply a waste of money.

Monetary cost. The monetary cost of scanners is important. Not because excessive cost should preclude the use of scanners, but because the question of whether that money could be better spent on other means of security. If airport scanners cost £X million, then the question: “Can we spend £X million on a different security measure that brings more security?” needs to be asked and answered. If the answer is yes, deploying airport scanners actually reduces our security.

The consultation document looks at the issue of privacy, but the analysis is superficial and in some cases erroneous. The following paragraphs look at some of the privacy issues.

The first thing to note about the privacy safeguards is that they reduce security. Paragraph 37 states:

“Security staff viewing images will be separate from, and not be able to identify, the person whose image they are viewing”

Security screening involves a combination of human and technological factors and works best when the two operates in conjunction. For example, when someone is searched, the unconscious signals given off by the person being searched aid the person doing the search. Airport scanners are rendered less effective if the operator cannot identify and observer the person being scanned. For example: “Do they look nervous?”, “Were they fidgeting in the queue?” and so on. By divorcing the human and technological factors in this way the code of practice makes the scanners less effective.

The decision not to offer those selected for scanning an alternative method of screening is flawed from both a privacy and security perspective. Paragraph 23 states: “For many people in society security scanners offer a less intrusive process than a hand search…”. Yes, but not for everybody. For some a hand search is less intrusive, and those who would prefer a hand search should be offered one. Privacy is a personal matter, and offering what the consultation asserts is a minority the alternative of a hand search would enhance privacy without compromising security.

If those who refuse to be scanned are simply not allowed to travel, then a large loophole has been introduced into the scanning process. A bomber can simply go to the airport and hope they aren’t scanned. If they do get selected for scanning then they can just refuse and they will be not allowed to travel – they can go home and try again another day. For this reason, anyone who refuses to be scanned should be searched. If we have compulsory hand search of anyone who refuses to be scanned, they we may as well allow all passengers the choice.

Paragraph 44 states: “The image produced does not show any distinguishing features such as hair or skin tone and it is not possible to recognise people from their facial features”. There are many features that people can be used to recognise people that will show up on the scanner image, for example height. If a queue of people in the airport contains an exceptionally tall or short person, then it will be possible to recognise that person from their image. Similarly if the queue near the scanner includes only one child. If the code of practice requires that an individual is not recognisable from their scanned image, then that requirement cannot consistently be met.

It must be noted that the some of the authority quoted is now outdated. In particular paragraph 19 states: “The Government believes that they[scanners] should be deployed as quickly as possible…” and paragraph 22 states: “The Government believes that this is a proportionate measure to maintain security levels.” The Government changed during the course of the consultation, so these statements may or may not be true.

Question 1: Do you agree with this approach? If not, what changes to the code of practice do you propose?

It is not clear what is meant by “approach”, since no overall approach has been defined, and indeed this is the first use of the word “approach” in the consultation paper. The overall approach to security has been a piecemeal reactive retrospective one. There was a shoe bomber, so we start to examine people’s shoes; there was a threat of a liquid bomb, so we ban liquids; there was an underpants bomber, so we decide to introduce technology that enables us to examine people’s underwear. Such an approach is doomed to failure, since it will fail to anticipate the next bomber.

The approach to airport security needs to be a holistic one. Rather than hastily introduce scanners, we should take a considered approach. We should have an overall review of airport security that looks at all aspects of security. This review should make recommendations about how security should be improved, and we should enact those recommendations.

Question 2: Do you agree that the safeguards outlined in the interim code of practice address all potential privacy concerns? If not, what else should be included?

No. As stated above, for both security and privacy reasons, people who do not wished to be scanned should be allowed to submit to an alternative form of search.

Question 3: Do you agree that the safeguards outlined in the interim code of practice satisfactorily address any potential data protection concerns? If not, what else should be included?

No. Images on the scanning machines may be photographed, so operators of the machines must not be allowed to have cameras or camera-phones on their persons. The issue that people can be identified by distinguishing physical characteristics (such as height) also needs to addressed.

Question 4: Do you agree that the safeguards outlined in the interim code of practice and HPA assessment satisfactorily address any potential health and safety concerns? If not, what further analysis would you wish the Government to undertake?

No. While I am persuaded that, when working correctly, airport scanners probably pose no undue health risk, I don’t think the possibility of malfunction has been sufficiently addressed. I suggest two further safety measures:
i) when a scan is made, a device inside the scanner records the actual level of radiation and displays and records this. This way, if the scanner becomes uncalibrated then it will be immediately apparent, and at most one passenger will be a
ii) Operators of scanners should wear a device that measures their accumulated exposure to radiation, and this is checked on a daily basis. This would help protect from accidental overexposure to radiation caused by, for example, radiation leakage from a scanner. Presumably operators of X-ray machines in airports already wear a similar device that is sensitive to X-rays; this would be a continuation of this practice.

Question 5: Do you agree that requiring airport operators to discuss with the DFT all prospective use of security scanners as outlined in the interim code of practice satisfactorily addresses the requirement for all equipment to undergo a suitable approval process? If not, what else should be included?

No. The requirement to “discuss with the DFT” is hardy a suitable approval process. An approval process should set out clearly the objective criteria for approval, and then should test applications for deployment of scanners against those criteria.

Question 6: Do you agree that requiring security officers operating security scanners to hold government security clearance and to have received training delivered in accordance with a DfT mandated security scanning training module before deployment satisfactorily addresses the issues of vetting and training? If not, what else should be included?

No. The main issue is vetting, not training. People who view scanned images need to be vetted over and above the standard vetting procedures.

Question 7: Do you agree that the requirements for keeping passengers informed outlined in the interim code of practice are sufficient? If not, what else should be included? And what additional means of communication do you suggest the Government or the travel industry should put into place?


Question 8: Do you agree that selection criteria defined in the interim code of practice provide an appropriate safeguard to ensure that passengers are selected for screening on a non-discriminatory basis? If not, how do you suggest passengers should be selected?


Question 9: Do you agree that the guidance provided in the Protocol section of the interim code of practice is satisfactory? If not, what else should be included?

No. The Protocol section states: “the details of the protocol are not published due to the security sensitive content.” Why even bother to ask this question?

Question 10: Are there any other issues that you would like to see the final code of practice consider? If so, what and why?

Yes, as stated above the issue of whether airport scanners actually work must be addressed. What tests have been conducted to show they work? The issue of their cost effectiveness must also be addressed. Are their other security measures that, for the same cost as the scanners, could deliver more security than scanners?

Yes. See preliminary section of my response.

A brief overview of COINS

Jun 12, 2010 7 comments

On June 4th, 2010 HM Government released Combined Online Information System (COINS) database, the main database used by HM Treasury budgeting. This is a significant milestone in the opening up of UK government data. The data is available here

This post gives a brief overview of that data, and links to sources that help understand that data. I’ve also done a blog post about a little programming project I did to recreate the government Public Expenditure Statistical Analyses(PESA) reports from the COINS data, see: Using the COINS data to recreate PESA reports

Overview of COINS data

The structure of the COINS data is explained in HM Treasury – Understanding the COINS data

Initially the government released four files in zip format. These unzipped to give:

  • fact_table_extract_2009_10.txt
  • fact_table_extract_2008_09.txt
  • adjustment_table_extract_2009_10.txt
  • adjustment_table_extract_2008_09.txt

A Very brief summary of the file structure

The released files contain the COINS data in unicode format, with fields separated by @ and one record on each line. All amounts are in thousands of pounds.

The Treasury maintains data for recorded spending (outturn), forecast spending (estimated outturn for the latest year) and planned spending (up to three years ahead). The COINS data also includes snapshots of the spending data. The Data_type field has values that reflect this, including “Outturn”, “Forecast Outturn ” (eg “Forecast Outturn March”), “Plans” and “Snapshot”.

A clear distinction is made between current and capital spending, based on Generally Accepted Accounting Practice (GAAP). The Resource_Capital field has a value of “Resource” or “Capital” to reflect this.

Departments are given firm three year spending limits called Departmental Expenditure Limits(DELs) within which they prioritise resources and plan ahead. Spending that cannot reasonably be subject to firm multi-year limits, or that relates to certain non-cash transactions, is included in Annually Managed Expenditure(AME). DEL and AME together make up Total Managed Expenditure(TME). The Budget_Boundary field has a value of DEL, AME or Not DEL/AME to reflect this.

Spending is by department. The Department_code field reflects this.

Useful links

Treasury Information

HM Treasury publish documents that may help to understand COINS and, and how that data may be used and aggregated. These include:

HM Treasury publishes a number of documents based COINS data. Public Spending Statistics are available here: Statistics on Public Finance and Spending.

Using the COINS data to recreate PESA reports

Jun 12, 2010 3 comments

On June 4th, 2010 HM Government released Combined Online Information System (COINS) database, the main database used by HM Treasury budgeting. I’ve written a brief overview of the data here. This post is about a little programming project I set myself to actually do something with this data.

The COINS data is in a raw format and needs to be converted to a more useful format before any useful reports can be generated. Mainly out of curiosity, I set myself a project to do something useful with this data. I wanted to do three things:

  • Extract useful subsets of the data that were small enough to be loaded into Excel.
  • Load the COINS data into an SQL database.
  • Recreate some of the government PESA(Public Expenditure Statistical Analyses) reports from the COINS data and check that my results are the same.

The project to do all this is written in python, and I’ve placed it on github here

Extracting useful subsets of the COINS data

The COINS fact_table_extract_2008_09.txt file contains 2,043,129 records that have a non-zero value field: of these 20,566 are outturn records, 1,610,900 are forecast records, 13,733 are plans records, and 397,930 are snapshot records. Older Excel spreadsheets are limited to 65535 rows, and I wanted to produce a useful data set that fit within this limit. The non-zero outturn records easily fit, I’ve made them available on rapidshare: coins2008_09 and coins2009_10.

Loading the COINS data into an SQL database.

I’ve extracted the 2008-09 and the 2009-10 COINS fact data into two SQLite databases: coins_2008_09_sqlite and coins_2009_10_sqlite

Checking my reports against published government data

HM Treasury produces Public Expenditure Statistical Analyses(PESA) reports. I wanted to recreate some of these reports from the COINS data.

I’ve produced outturn Resource, Capital, DEL and AME reports by departmental group from the 2008-09 COINS data, and checked it against the estimated 2008-09 outturn from the PESA 2009 report and the actual 2008-09 outturn from the PESA 2010 report.

The PESA data used was:

The spreadsheet PESA 2010, Chapter 1 tables which contains tables of DEL, AME and TME grouped by Resource and Capital expenditure (from webpage PESA 2010 section 1 – Budgets). 2008-09 subtotals by departmental group are provided in Column G (2008-09 outturn) of the worksheets:

  • Table 1.3 Resource budgets, 2004-05 to 2008-09 (by departmental group)
  • Table 1.6 Capital budgets, 2004-05 to 2008-09 (by departmental group)

The spreadsheet PESA 2010, Chapter 1 tables which contains tables of DEL, AME and TME grouped by Resource and Capital expenditure (from webpage Public Expenditure Statistical Analyses 2009). 2008-09 subtotals by departmental group are provided in Column G (2008-09 estimated outturn) of the worksheets:

  • Table 1.5 Resource budgets, 2003-04 to 2010-10
  • Table 1.10 Capital budgets, 2003-04 to 2010-11

I produced my report by running the python program against my generated 2008-09 SQLite database. (See my readme file for a full description of how to recreate these reports.)

The comparisons

The following tables give my results and the PESA result side by side for comparison.

Resource DEL by departmental group (millions)
Children, Schools and Families 46,848 46,848 46,848
Health 118,899 90,278 92,455
Transport 6,245 5,083 6,546
Innovation, Universities and Skills N/A N/A 16,666
CLG Communities and Local Government 28,755 4,105
Home Office 9,198 9,198 8,926
Justice 9,234 9,235 9,283
Law Officers’ Departments 723 724 733
Defence 29,790 32,620 37,889
Foreign and Commonwealth Office 2,027 2,027 2,025
International Development 4,758 4,758 4,835
Energy and Climate Change 292 293 1,016
Business, Enterprise and Regularity Reform 175 N/A 1,594
Environment, Food and Rural Affairs 2,449 2,446 2,654
Culture, Media and Sport 1,455 1,456 1,633
Work and Pensions 7,937 7,937 8,059
Scotland 24,090 24,090 24,599
Wales 12,775 12,799 12,970
Northern Ireland Executive 7,926 7,926 8,117
Northern Ireland Office 1,177 1,177 1,342
Chancellor’s Departments 4,463 4,473 4,826
Cabinet Office 413 1,995 2,049
Independent Bodies 586 791 806
Resource AME by departmental group (millions)
Children, Schools and Families -2 10,652 10,914
Health 2,870 14,984 13,934
Transport 602 603 3,878
Innovation, Universities and Skills N/A N/A 274
CLG Communities and Local Government 1,282 621
Home Office 709 710 363
Justice 307 439 725
Law Officers’ Departments 9 9 N/A
Defence 392 6,193 6,864
Foreign and Commonwealth Office -28 -28 -7
International Development 145 213 397
Energy and Climate Change 2,403 2,403 4,589
Business, Enterprise and Regularity Reform 1 N/A 819
Environment, Food and Rural Affairs 0 0 -52
Culture, Media and Sport 3,890 3,890 3,911
Work and Pensions 135,344 135,344 135,546
Scotland 4,456 2,495 3,129
Wales 110 138 520
Northern Ireland Executive 6,711 6,104 8,142
Northern Ireland Office 395 396 264
Chancellor’s Departments 71,836 71,209 80,757
Cabinet Office 1 7,174 7,221
Independent Bodies 9 15 24
Capital DEL by departmental group (millions)
Children, Schools and Families 5,519 5,519 5,634
Health 11,602 4,370 4,561
Transport 7,182 7,252 7,283
Innovation, Universities and Skills 0 N/A 2,123
CLG Communities and Local Government 7,233 7,112
Home Office 836 836 862
Justice 912 912 975
Law Officers’ Departments 8 9 12
Defence 7,754 8,980 8,604
Foreign and Commonwealth Office 226 227 218
International Development 875 875 891
Energy and Climate Change 1,667 1,667 1,688
Business, Enterprise and Regularity Reform 2 N/A 16
Environment, Food and Rural Affairs 609 610 618
Culture, Media and Sport 841 842 791
Work and Pensions 91 91 57
Scotland 3,332 3,333 3,337
Wales 1,626 1,627 1,656
Northern Ireland Executive 1,233 1,233 1,137
Northern Ireland Office 67 68 78
Chancellor’s Departments 282 282 293
Cabinet Office 41 397 416
Independent Bodies 32 37 42
Capital AME by departmental group (millions)
Children, Schools and Families 0 N/A N/A
Health 3,780 14 14
Transport 22 N/A N/A
Innovation, Universities and Skills 0 N/A 4,230
CLG Communities and Local Government 3,766 516 -
Home Office 0 N/A N/A
Justice 27 N/A N/A
Law Officers’ Departments 0 N/A N/A
Defence 85 N/A -
Foreign and Commonwealth Office 41 N/A 0
International Development 86 N/A N/A
Energy and Climate Change -35 -279 -419
Business, Enterprise and Regularity Reform -210 N/A N/A
Environment, Food and Rural Affairs -12 1 0
Culture, Media and Sport 571 572 717
Work and Pensions 126 136 142
Scotland 900 180 187
Wales 293 168 177
Northern Ireland Executive 611 378 413
Northern Ireland Office 0 N/A N/A
Chancellor’s Departments 86,001 85,822 85,590
Cabinet Office 0 N/A N/A
Independent Bodies 0 - -

The differences

There is good agreement between some of my calculated results and the PESA reports, but there are also some significant differences between the figures.

In the “Resource DEL” and “Capital DEL” tables there is generally good agreement, except for Health, Transport, Defence, Cabinet Office and Independent Bodies departmental groups. The Cabinet Office and Independent Bodies discrepancies could be because some departmental data is missing from the COINS data (see next section). I think the differences in the values for Health are probably because the PESA reports do not include the figures for Scotland, but I haven’t checked this. I have no explanation for the differences between the figures for Transport and Defence.

The agreement for the “Resource AME” and “Capital AME” tables is less good. Some of this disagreement is for the same reasons as the “DEL” differences, but I suspect the main reason for the differences is that my SQL query for the “AME” figures is not correct. (Any corrections to my SQL queries are most appreciated, as are any bug reports or corrections to my programs.)

Missing COINS departmental data

The COINS data does not seem to have spending figures for the following government departments:

  • Assets Recovery Agency
  • Department for Business, Enterprise and Regulatory Reform
  • Office of Communications
  • Forrestry Commision
  • National Investment and Loans Office
  • Royal Mint
  • Office of Government Commerce
  • Security and Intelligence Agencies

This may because the departments have been renamed, or subsumed into other departments.

Next steps

I’ve satisfied my curiosity about the COINS data. I’ve managed to recreate some of the PESA reports, and although there are differences between my reports and the PESA reports, the differences are mostly explainable. I’m not Cliff Stoll, and won’t be tracking down the exact nature of the differences. I don’t plan to do any further work on the COINS data, but may look at future data that is released, especially if it is more detailed.

Feel free to use my programs if you wish – it would nice if you credit me if you do so. I’m also interested in other programming projects that use the COINS data – feel free to post a comment with a link if you do anything interesting with the COINS data.

The Myth of Panic

Apr 8, 2010 Leave a comment

We see it time and again in the movies. A disaster is about to happen: a nuclear bomb is about to explode in a major city, a deadly virus has been released from a laboratory, a giant meteor is about to strike the earth, an alien spaceship is spotted heading towards the earth. And the reaction from the president is always the same: “We must not tell the public or there will be mass panic and hysteria”.

Except there won’t.

The assumption of public panic is a useful plot device for disaster movies, but as an instrument for policy formulation it is defective. Policy makers seem to assume that the public will panic, but research shows the opposite. Baruch Fischhoff, professor of social and decision sciences at Carnegie Mellon University and president of the Society for Risk Analysis, says “people, however stressed, almost always keep their wits and elevate their humanity”. In his article A Hero in Every Aisle Seat he says:

Studies of civilians’ intense experiences in the London Blitz; the cities of Japan and Germany in World War II; the 1947 smallpox outbreak in New York; the earthquake in Kobe, Japan, in 1995; and even fires have found that people, however stressed, almost always keep their wits and elevate their humanity.

In his article Panic: myth or reality, Lee Clarke, Associate Professor, Department of Sociology, Rutgers University states:

Before, during and after disasters, the ‘general public’ warrants trust and respect. Panic is often used as a justification by high-level decision makers to deny knowledge and access to the public, on the presumption that people cannot handle bad news. Research on how people respond to life threatening disasters and the stories form the World Trade Center show that people handle even the most terrifying news civilly and cooperatively. Our leaders would do well to see us as partners in recovery rather than a ‘constituency’ to be handled.

In The Swine Flu Panic That Wasn’t. Mass hysteria fails to materialize. Again. Jesse Walker says:

People are sharing information, they’re seeking out information, they’re asking questions about whether or not they have the symptoms,” says Jeannette Sutton, a researcher at the Natural Hazards Center at the University of Colorado at Boulder. “Those are not incidents of panic or hysteria. That’s rational thinking, where people are asking questions and trying to make decisions based on the information they have available to them.”

It’s not as though there haven’t been any destructive overreactions to the H1N1 flu. It’s just that they’ve come from officials, not the general public.

A John Hopkins University study dispels panic myth and suggests ways to involve the public in response to a bioterrorist attack.

Planners and policy makers have long discounted the public’s ability to participate in a response to bioterrorism, because of a belief that an attack would create mass panic and social disorder.

However, researchers at the Johns Hopkins Bloomberg School of Public Health, who reviewed the public’s response to the terrorist attacks of September 11th, the recent anthrax mailings, and other disasters concluded that the public does not react with panic but with effective and adaptive action and can be an valuable response force and that should be considered in biodefense planning.

It is a myth that a community’s first response to a crisis is panic.

The public are much smarter than portrayed – we pay much more attention to the actual reported numbers than the apocalyptic predictions of the media and politicians.

The assumption of public panic results in poor policy decisions. It’s not the public who panics, it’s the politicians.

The form of jury-less criminal trials

Apr 1, 2010 Leave a comment

The first major criminal trial to be held without a jury in Britain for more than 350 years (the Heathrow Airport Robbery Trial) concluded yesterday. Until this case, and since the abolition of the Court of Star Chamber in 1641, all serious criminal offences on indictment were tried by jury. Trial by jury traces its roots back to Article 39 of the Magna Carta signed by King John in 1215.

Trials on indictment without a jury were made possible by The Criminal Justice Act 2003. A judge may now order a jury-less trial in the specific cases of complex fraud and jury tampering.

There has been much discussion about the ramifications of the removal of the right to trial by jury in criminal trials. I think that the right to trial by jury of ones peers is an important right, and I think it was wrong to remove that right.

But, in the context of our situation where we have had our right to trial by jury removed, I want to talk about something else – the form of those jury-less trials.

The jury should be replaced, not eliminated

In a jury trial, broadly speaking, the judge rules on matters of law and is responsible for sentencing; the jury decides matters of fact (by evaluating the evidence). What’s more the jury consists of individuals with varying opinions and backgrounds – the debate that occurs between jurors in the jury room is an essential part of establishing a verdict. A judge sitting on their own has nobody to challenge their assumptions.

When a case is tried by a single judge, both the form of the trial and the dynamic of the courtroom is radically changed.

Part 7 (Trials on indictment without a jury) of The Criminal Justice Act 2003 addresses the problems of complex and lengthy trials and the problems of jury tampering. These are problems resulting from having citizen jurors, not problems resulting from a having a jury per se.

Rather than solving the problems by complete removal of the jury, I believe that the less drastic measure of replacing the citizen jury by an appointed jury should be adopted. In particular:

  1. The jury should be replaced, not combined with the role of judge. That is there should be a judge who acts as judge in the case, and a separate judge who sits unrobed in the jury box and acts as proxy for the jury. Being unrobed is important, since it is a constant reminder to the court that the judge is acting as jury, not as judge and jury.
  2. Having said (2), and mentioned the importance of debate within the jury, I think that if the jury is replaced, then it should actually be replaced by at least two judges who sit unrobed in the jury box.
  3. The sitting judge should treat the appointed jury of judges just like a citizen jury.
  4. Having three judges for a trial is undoubtedly expensive, but the expense serves as a disincentive to removing a citizen jury.


Judge with no jury convicts Heathrow robbery gang

Armed raiders jailed after trial without jury

Heathrow robbers guilty: the trial without jury that made history

Four jailed for £1.75m Heathrow robbery

Law Review: No angry men

No angry men: first trial without jury begins

Tags: , , ,

Get every new post delivered to your Inbox.