Click here for PDF version
Review by Walter Forrest

Damned Lies and Statistics: Untangling Numbers from the Media, Politicians, and Activists
By Joel Best
University of California Press, 2001, 190pp,
$US19.95, ISBN 0520219783

The main contention of this book is that statistics Ôare products of social activityÕ and as such are susceptible to errors. Instead of assuming that statistics are facts that simply exist, readers are cautioned to be vigilant in their acceptance of their use. ÔTo sort out the good statistics from the badÕ, Best counsels his readers to think about three things every time they encounter a new statistic: who created it, why was it created, and how was it created? The purpose of this book is to help readers make sense of their answers to those questions in order to develop a more critical approach to the interpretation of statistics.

The book begins with a brief introduction to the rise of social statistics and their uses in the construction of social problems. The book then turns to a discussion of Ôthe most common problemsÕ concerning the creation and interpretation of statistical data. These concern the creation of spurious numbers based on poor definitions (for example, false negatives and false positives), erroneous and sometimes fraudulent estimates, the context and wording of questions in public opinion polls, and sampling error.

Of course, even accurate statistics can be made erroneous as people interpret and relay information incorrectly, often unintentionally. To that end, in one of the more interesting passages of the book, Best deals with the topic of mutant statistics and describes the ways in which these numbers are created. These occur by drawing inappropriate generalisations from a statistic; Ôtaking a number that means one thing and interpreting it to mean something differentÕ, confusing the meanings of more complicated statistics, and compounding errors in subsequent mutations.

The problem of mutant statistics is neatly illustrated in the book by the following example. An article published in a scholarly journal claimed that Ôevery year since 1950, the number of American children gunned down has doubledÕ. According to this statistic, even if one child had been Ôgunned downÕ in 1950 the number killed in the year the article was published would have been 35 trillion. The origin of this mutant statistics was the much less spurious claim that Ôthe number of American children killed each year by guns has doubled since 1950Õ. Simply by trying to repeat the original figure, the anonymous author had fundamentally transformed its meaning.

Chapter five looks at some debates over statisticsÑincluding a timely review of US debates on the collection of the Census. The book closes by making a case for a more critical approach to the review and interpretation of numbers based on an appreciation of Ôthe inevitable limitations that affect all statisticsÕ.

For a book that argues against the use of spurious evidence to advance arguments, Damned Lies and Statistics seems to wage a few arguments without evidence of any kind. From the outset, Joel Best implies that the use of numbers as social tools encourages their deliberate misuseÑespecially by ÔactivistsÕ who wish to draw attention to social problems.

At one point, the author argues that when forced to make estimates concerning the numerical extent of social problems (for example, the number of homeless people, the number of non-reported crimes), activists will err on the side of exaggeration largely because they have an incentive to do so. In another section, it is claimed that in constructing statistical definitions, those wishing to draw attention to social problems are also more inclined to support broader definitions that exaggerate a problem (and are therefore less concerned with false positives than false negatives).

All of these claims, including the overall argument regarding deliberate falsehoods, are made without much supporting evidence. To me, this detracts from the book, especially because the argument is not at all necessary. It would have been sufficient for the author to flag the potential sources of error, noting some examples of those errors, rather than claim that there is a systematic tendency by some people to generate certain types of error. After all, in the interests of a critical approach, readers should be somewhat sceptical of all statistics.

Those criticisms aside, the book provides a good overview of the principal flaws in the development of descriptive statistics in a very readable and non-technical fashion. It is filled with numerous examples of bad statisticsÑthough not many examples of good statisticsÑmaking it a clear guide to what not to do. For these reasons, it is probably most relevant to those people who frequently encounter statistical evidence, but who feel unable to evaluate its veracity.

By contrast, this book is not likely to interest those readers who have either studied statistics at some point in their careersÑas distinct from attending a course on statisticsÑor for whom statistical scepticism is commonsense, although many of the examples are entertaining. The sources of errors reviewed in the book are not new and are covered by most introductory courses and textbooks on the subject.

Moreover, the book deals only with Ôthe sorts of statistics typically addressed in the first week or so of an introductory statistics courseÕ. The book is also very repetitive and could easily have made its case more succinctly. I would have preferred it if, in the space saved by more rigorous editing, the author had extended his gaze beyond descriptive statistics to look at the errors involved in inferential statistics. After all, it is usually with respect to inferential statistics, in which claims are made about correlation and causal relationships, that the lies are most damned.

Review by Richard Grant

Social Policy, Public Policy: From Problem to Practice
by Meredith Edwards with Cosmo Howard and Robin Miller
Allen & Unwin, 2001, 232 pp, $35, ISBN 1 86448948 0

Social Policy, Public Policy is a book written by a policymaker for policymakers. Meredith Edwards, a distinguished public servant in various social policy portfolios, has compiled four case studies designed to show the policy process Ôfrom problem to practiceÕ.

The case studies are the Youth Allowance (AUSTUDY), the Child Support Scheme, the Higher Education Contribution Scheme and the Working Nation programmes. Edwards chooses examples that led to long-term policy breakthroughs. Not surprisingly, she emphasises the merit in pursuing the problems that underlay these schemes and the actual success of bureaucracy in delivering them. Her intent is clearly to affirm the good that government can do and more particularly, the pioneering efforts of Australian administrators in social policy. There are strong grounds for these claims. A HECS style system, for instance, is unique in the higher education sector.

The reader will note more than a tinge of partisanship in the book. All four case studies are chosen from LaborÕs time in office with contrasts drawn to the approach of a Liberal administration. The chapter on higher education is titled HECS and not Fees. The chapter on the Working Nation programmes employs an analystÕs quote as Ôa very valuable social experiment which was aborted for political reasonsÕ (174). In the chapter on the Child Support Scheme, Edwards underlines the governmentÕs eagerness to Ômake its mark in moulding public attitudes towards quality family lifeÕ (58). There is undoubtedly a strong theme throughout the book that the reforms were pursued, and were successful, in part because of their ideological character.

EdwardsÕs main focus, though, is on the mindset and method of policymakers. Each of the four reforms are presented under the various headings of Ôhistory and contextÕ, Ôputting the problem on the agendaÕ, Ôdata and researchÕ, Ôdeveloping optionsÕ, ÔconsultationÕ, ÔpublicityÕ and ÔevaluationÕ. The nexus of policy and politics is captured well through this structure. The ability to clearly identify both the conflict and progression of elite thought on all four of these complex policy areas is a credit to both this method and the level of detail.

Unfortunately, the book lacks the theoretical observations that would attract a wider audience and give the case studies greater depth. Edwards does not attempt to weave the first chapterÕs rather thin consideration of normative policymaking issues into the case studies. While the accent on the players and the process is quite appealing, it is at the expense of the policymaking literature. The work of notable writersÑLindblom, Wildavsky, North, March and OlsenÑis not considered in light of the case study findings. On the evidence Edwards presents, there is cause to question LindblomÕs famous characterisation of Ôthe science of muddling throughÕ. The impression given from the four case studies is of a high level of bureaucratic structure and coordination.

One of the main conclusions of the book is how policymakers have successfully balanced equity and efficiency concerns. There are numerous references to the need for savings. The higher education tax is pitched in terms of redressing Ômiddle class welfareÕ given that: ÔNot only had they not benefited personally, but the majority of Australian taxpayers were, on average, less well off than those they were supporting through the provision of ÔfreeÕ tertiary educationÕ (111). The impetus for the child maintenance payment was the need to reduce sole parent pension outlays.

Edwards recognises that the HECS and child support initiatives were easily saleable to the taxpayer. The funding of labour market programmes in Working Nation was both more expensive and more contentious in terms of its effectiveness. Edwards describes these programs in terms of helping the long-term unemployed back into work Ôwithin the framework of fiscal responsibilityÕ (181). Perhaps because of the difficulty in ascertaining their effectiveness, this spending on labour market programmes has been a point of major political contention. In terms of the inputs into the policy process, Edwards makes much of the contribution of academics. While this is well placed, it is at the expense of the important contribution of the economic portfolios in impressing the need for savings. Given EdwardsÕ characterisation of these social policies as enduring and innovative achievements, the role of Treasury and Finance in creating the competitive framework for this innovation should have received more attention. Such arguments are important as a rejoinder to Michael PuseyÕs Economic Rationalism in Canberra. Undoubtedly, the economic portfolios assumed great power and influence during the Hawke/Keating years, but as Edwards shows, this was largely to the improvement, not the detriment of equity.

Overall, Edwards and her co-editors are to be commended for an original text. Books on bureaucratic process are not known for their appeal, but this work should prove useful for those interested in some of the major public policy reforms of the past 20 years. A rewrite might consider more attention to either the public policymaking literature or the actual conceptualisation of the policy reforms. Of course, this depends on the intended audience.

Review by Michael Warby

Anti-Liberalism 2000: The Rise of New Millennium Collectivism
by David Henderson
Institute of Economic Affairs, London, 56pp, £7.50, ISBN 0 255 36497 0

I have always enjoyed David HendersonÕs company, both in person and in his writing, and the recent publication by the Institute for Economic Affairs of his 2000 Wincott Lecture, Anti-Liberalism 2000, provides an opportunity for us all to enjoy his company all over again.

In the lecture, Henderson once more revisits ideas familiar to us from his other writings: that claims about the dominance of economically liberal opinion are greatly overdone, that the influence of economists is easily exaggerated, that the course of events cannot be explained by a narrow interest-group view of public debate, that the persistence of what he calls Ôpre-economicÕ ideas (or Ôdo-it-yourself economicsÕ) is very important in explaining the pattern of opinion, that economically liberal reforms have often come from governments of the centre-left, that the scope of such reforms is easily exaggerated, that the tide of opinion moves back and forth. He also sets out various propositions which he believes are typical of anti-liberal economic opinion.

Henderson points out that while there was a deterioration in general economic performance (particularly unemployment) from the early 1970s, this cannot be blamed on economic liberalisation since that did not really get underway until the early 1980s. He also absolves liberalisation from (amongst other things) leading to greater volatility, noting that economic growth performance across the OECD as a whole is more stable after liberalisation gets underway than in the two decades beforehand.

What Henderson seeks to delineate is the coming together of interests and perceptions. In particular, the propagation of two propositions he judges to be characteristic of what he calls new millennium collectivism. The propositions are (i) Ôa market economy, even a well-performing one, is heavily populated with non-beneficiaries and victimsÕ and (ii) Ôtheir well-being depends on deliverance from aboveÕ.

While he notes that the vision is not new, he identifies some new aspectsÑthe rise of NGOs (non-government organisations), the widening circle of victims, the rise of new interventions in the labour market (regulation for equal opportunity, anti-discrimination, human rights and affirmative action in the workplace), the scare over globalisation.

He further identifies a common presumption in all thisÑthe presumption of injusticeÑand the scale of the aspirations new millennium collectivism brings forthÑnothing less than a desire to regulate the globe. Though his main concern is to delineate these ideas and associated interests, Henderson does provide some critique of these ideas and considers the state of opinion within the economics profession itself.

Geoff HarcourtÕs comments on HendersonÕs lecture, which are included in the IEA publication, would have undoubtedly been much more interesting if they concentrated on being comments on what Henderson said. Unfortunately, much of HarcourtÕs comments are taken up by clarifying his own position (by my count, in the space of 2,100 words he uses the world ÔIÕ 60 times, ÔmeÕ 7 times and ÔmyÕ 5 timesÑthat is, there is some reference to himself about once every 30 words or so).

Alas, much of what passes for public debate in the various Anglomorph countries (as the late Frank Knofelmacher called them) is precisely aimed at showing that the writer/speaker is a virtuous sort of person. It is perhaps the most characteristic achievement of the baby-boomer generation that so many have turned public debate and much of what passes for scholarship from a public good (an attempt to find out the truth and to worry at what makes for a good society and a good life) into a private good (use of writing and speaking as moral display).

This may be the most powerful single impetus in generating the opinion conformity which is particularly stultifying in Australia but which has baneful effects right around the Anglomorph world (The Shadow University: The Betrayal of Liberty on AmericaÕs Campuses by Alan Kors and Harvey Silvergate documents just how virulent and pernicious the attempt to restrict legitimacy in debate to being a prerogative of the members of the club of the virtuous can be).

The ÔClub VirtueÕ approach to debate is particularly useful for anti-liberalism, because so much of that is concerned with good intentionsÑperfect for moral display. Much of anti-liberalism holds that the effect of policy (or, at least, the policy they agree with) can be inferred from its intentions. Economic liberals, on the other hand, tend to be concerned about consequences, a far more grubby businessÑand one more complex to talk about than simple intentions, a significant PR disadvantage in the age of the 20 second grab.

Another aspect of the culture of virtue is that status is a positional good, which creates a constant demand to have the status of being in the moral vanguard. An ever-widening circle of victims for people to be seen to be concerned about is a natural response to this ever-present demand.

The word Ôneo-liberalÕ is a classic in the virtue genre, being a term of insult rather than analysis. The term has the great advantage of allowing the writer or speaker to signal that they, of course, do not share the Ôneo-liberalsÕ (sic) pernicious ideas. David Henderson is a world away from such intellectual perversions. He concludes his comments on Geoff HarcourtÕs comments on a note that any genuinely inquiring intelligence can surely agree with, at least regarding what philosophers would call contingent propositions: Ôall our doctrines and ways of thinking are in a sense provisional: they have to meet the test of experience in a world that is both highly complex and subject to unceasing and often unforeseen changesÕ.

As long as the classical liberal cause can bring forth such ornaments as the clear-thinking and writing, yet subtle and sophisticated, mind of David Henderson, it will have intellectual riches indeed to draw upon. Anti-Liberalism 2000 is another fine product of that mind for us to enjoy.

previous

next

Policy is the quarterly review of The Centre for Independent Studies. For more information on subscribing to Policy, click HERE

If you are interested in the Centre's activities and publications, why not subscribe to e-PreCIS, our regular email update on the latest news and events.

(e-PreCIS requires html capable email facilities, such as Microsoft Outlook Express or Netscape Messenger)