It’s time once again for your comments on any topic (civilised discussion and no coarse language please).
Suggested discussion starter: Is there an optimum population for Australia, and if so, what is it?
17 thoughts on “Monday Message Board”
Comments are closed.
John Quiggin 
In “Polical Religion” you stated that that Professor Mark Kleiman effectively demolished the worth of faith based schemes ( my words). He was commenting on a scheme by InnerChange Freedom Initiative (IFI)which is headed by Chuck Colson.
After a length of time I have viewed both the original study ( http://www.crrucs.org/8_research_pdf/innerchange_freedom.pdf) and Professor Kleiman’s critique.
I believe he has demolished his reputation.
The study, conducted by Byron Johnson of the University of Pennsylvania, compared 177 inmates who had participated in IFI with a similar group, called the control group, that did not participate. Itâs important to note that the IFI group included both those who graduated from the program and those who dropped out or whom the state moved out.
Johnson found that when you looked at all IFI participants, including dropouts, the re-arrest and reincarceration rates were essentially the same as the prisoners who didnât participate÷understandable. But the difference IFI makes comes out when you looked at those who completed the program: They were two times less likely to be re-arrested, and the study showed an 8 percent recidivism rate compared with 20 percent in the control group, or 67 percent in the country at large.
How Kleiman can assert people who drop out of a program for what any reason should be counted is beyond me except if you wish to use selective statistics.
Surely looking at the effectiveness of a program you look at the prople who complete the prorgam.
Whatâs more, if you applied Kleimanâs standards across the board, just think what would happen to the literacy programs that he cites as his favorite example. If we included the post-release record of those who attended just a little bit, but never learned to read, what kind of result would you get?
As IFI say themselves…
IFI isnât a magic wand. We have never argued that mere exposure to the program would make a lasting difference. Itâs a program! The goal is to “re-educate” men over the course of a year, men who have spent most of their lives breaking the law, and we want to make them think and then live in a totally different way. IFI seeks to undo the habits of a lifetime. When you see it in that context, the post-release success of our graduates ought to encourage us and thrill the country, not prompt accusations of “fudging and fraud.”
This leads to to think that Professor Kleiman just had little understanding of what IFI’s endeavours were.
I know what my old statistics lecturer would have said about Professor Kleiman’s review and it wouldn’t have been all that favourable.
I believe Phil Ruthven has in the past predicted 80m or so, but haven’t looked at him for a long time.
I understand he makes reasonably realistic assumptions about politico-economic organization, international order and connectedness, technical progress, etc.
Do you have a model in mind John?
How long is a piece of string?
If one is willing to put up with North Arfican environmental values, East Asian urban values and South Asian family values then the sky is the limit. It is assumed that there are non-ecological constraints on increasing population:
Economic: balance of payments on primary goods
Cutural: integrative-proneness of migrants
An optimum population is fundamentally dependent on the sustainability of the ecology under conditions of greater use.
To support a larger population there must be an increased Carrying Capacity of the Land (CCL), given that, as Mark Twain would put it, “they ain’t makin any more of it”.
The CCL can be increased by increasing:
extensity of land available for use
intensity of land currently under use
The CCL is a function of water availability, soil fertility and air purity – all of which are under strain in Australia. Since, pace the Murray, we are already over-working the marginallly productive land this rules out increasing CCL extensity.
This leaves increasing land-use intensity, by more efficient land use ie using conservation, fertilisation, restoration to support more people per given unit of land/water. If this can be done, then the population can be increased, but only if the migrants are:
economicly productive
culturally integrative
This rules out factory-fodder and fundies.
Israel (inner) is the only example of a post-industrial democracies that has allowed massive population increases without experiencing social upheaval.
Japan is the best example of increased land intensity usage without incurring a massive trade balance deficit.
Saudi Arabia has managed to utilise water resources to cope with a larger population in an arid climate.
I conclude that Australia may increase it’s population by following:
Israel’s cultural policies
Japan’s economic policies
Saudi Arabia’s ecological policies
This is unlikely, so I conclude that it will not happen unless Australia experiences a real ecological crisis whilst needing more population.
A simpler solution to a stagnant population would be to just increase domestic population growth by:
increasing the birth rate of educated women
reducing the emigration rate of motivated men
What matters most is not the aggregate size of the population, but it’s per capita productivity.
The concept of optimum population provides a useful perspective upon, and even potentially a framework for, long-term economic policy formation.
In conjunction with the traditional aggregate measures such as BoP, etc., as Jack has suggested above, wouldn’t broader social utility measurement tools, such as the so-called ‘quality of life’ indices being developed, be more appropriate inputs to attempts at constructing optimum population models?
I raise this because Jack’s comments evoke the awful fantasy that an ultimate objective imposed on optimum population policy could be to sustainably pack together as many humans as possible into the available space, while the policymakers presumably would externalize their personal quality-of-life costs of this by retreating to country houses.
So the requirement of sustainability is insufficient. A consensus level of quality-of-life, for a diverse range of species, including humans, is also an input that would required by any civilized model.
The other burning question optimum population considerations raise is, who decides? One of the glaring weaknesses in our system of government (see also comment 9 in the ‘Corruption’ blog) is election cycles that make it rare for representatives to be able to effectively address issues that require long-term policies.
So, Pusey-style, there is the danger that they might be led by the nose by the technocrats (of whatever origin) on this one. American technocrats are already threatening, in the nicest possible way, to get the WTO to force us to lift the verious moratoriums currently in place on the import and commercial use of genetically-modified agricultural products.
(By the way, Phil Ruthven is at http://www.ibisworld.com.au)
The last bit of Jack Strocchi’s post nails it, in terms of a logical, low-impact solution to the looming ageing population (and therefore shortage of taxpayers) crisis.
Alas though, I think that is has bugger-all chances of being seriously thought about, much less implemented. The whole “Population Summit” thing was yet another cynical, baby boomer attempt to derail constructive discussion about a real problem.
Until boomers wake-up to the huge economic and social faultline they have created, debate in this country is going to get progressively more pointless.
as the posts above suggest, a country’s optimum population size is a function of its national cultural, social and economic values – and as these change over time, so does the notion of a desirable population.
I find myself disagreeing with some of what Jack posts, alas. The argument about carrying capacity of the land has never appealed; yes, we are putting strains on our water and land resources, but that is driven by agriculture, and that in turn by exports. It’s not Australia’s own population that is the source of the overuse of the resource; we export most of our agricultural product, ie the transformeed water, soil etc. Nonetheless it’s a common misperception. Thus I find myself getting annoyed when I hear farmers’ representatives claiming that if there were better use made made of water or restrictions on land clearing city folk wouldn’t have food to eat!! (farmers’ incomes might decline, and that’s a valid point, but entirely a different one).
On the point about dealing with stagnating population by “increasing the birth rate of educated women” – the difficulty is that nowhere in the Western world has there been any policy that has succeeded in doing this. Declining fertility is a constant, whether in Catholic or non-Catholic countries, whether in countries with financial incentives for childbirth or without them. It appears the best approach is to adjust to declining birthrates rather than fight against the incoming tide. There is some assistance to be gained from immigration, but not enough to offset the impact of declining fertility (for my own thoughts on this may I be immodest enough to refer to the following – http://www.abc.net.au/rn/talks/perspective/stories/s904102.htm.)
stephen
Homer is not correct in his analysis of the IFI program (the first comment posted on the Monday message Board).
If the IFI program had the same average rate of recidivism as the control group (20 per cent), then those that dropped out of the IFI program would have to have had a worse rate of recidivism than the average for the control group (ie. more than 20 per cent). This leads to two alternative conclusions:
(1) The IFI program helped some inmates (by reducing the recidivism rate to 8 per cent) but had the opposite effect on others (who dropped out and eventually re-offended). That is, for some inmates the program had an adverse effect.
(2) More plausibly, the control group also could be divided into two groups – those that would have completed if they had done the course and those that would have dropped out. The rates of recidivism in the first sub-group of the control group could be 8 per cent (the same as the equivalent program participants)and the rates in the second sub-group could be more than 20 per cent. In other words there is no evidence of a beneficial effect. More data is needed to adequately answer this question. You would have to know in advance what characteristics make it most likely for someone to drop out of the course and statistically control for these. (However, if you had that information, you could use it to “cream skim” particpants into the program in which case the average outcome for the IFI program would need to be significantly higher than the control group before you could accept that it has a beneficial effect.)
So what Kleiman and Quiggin are saying still stands.
Homer –
Ever heard of selection bias?
Of course you have to include the dropouts. The group who complete the IFI course will largely be those compliant individuals who were least likely to reoffend anyway. If you want to know what the actual difference the course made, you have to correct both for selection bias on entry to the course (using either a randomised experiment or some pretty sophisticated microeconometrics) and for dropout bias.
Reading Tim Flannery’s ‘Beautiful Lies’ in the Quarterly Essay, issue 9, and recommend it on many levels.
Some of the issues raised have been debated (with feeling) in the media recently, but the essay as a whole is far more interesting than these talking points.
Mark and DD,
How can you include people who didn’t complete the program when evaluating the program?
do you think JQ would like to be evaluated by his Uni by all students that enrolled in his courses ( assuming he conducts courses) or merely those who sat for examinations?
By the way why has Prpfessor Kleiman used different methods to evaluate education programs he favours. Indeed the very methods he criticizes in this study.
stephen in a comment above says that we “export most of our agricultural product”. This rang some bells with me. It has taken me a while to nail it, but I think I’ve done it.
In June 2001 I recall some controversy on this issue as the result of work done by Mark McGovern, an academic in the School of Marketing and International Business. My memory tells me that he reckoned the amount exported was between 22% and 36% of production, and probably closer to the lower end. The reason for the band and the doubt was related to how much product ended up in exported manufactures.
McGovern’s paper ‘On the unimportance of exports to Australian agriculture’ appears to have been published in the Australasian Journal of Regional Studies, v5, n2, 1999 pp 229-252.
As far as I can make out it became controversial in June 2001 as a result of an article by Colin Teese in “News Weekly” published by the National Civic Council. In a subsequent letter McGovern states that only 7 of 53 sectors export more than 50% – ginned cotton, wheat, wool, wool scouring, fresh meat and lobster/crayfish. You will note that 3 of these are not food.
In the 16 June issue Prof Emeritus Rod Jensen wrote in saying that McGovern’s figures are basically right, but that ABARE, farmers’ groups etc basically did not want to know.
McGovern’s figures apparently showed that at the farm gate $5 billion of $23.5 billion was exported. The reference year was 1993/94 but I think he had a look also at 1996.
Since then the paper and its implications have disappeared from the face of the earth. Richard Sanders, the ecological economist and erstwhile anti MAI campaigner referred to it in his submission to the senate on GATS/USFTA. He said it had some minor methodological flaws, but was basically OK.
The implication of this, if McGovern was and is still right, is that we may have the farm lobby tail wagging the trade policy dog, but I’m just an interested layman.
I’d be interested in any comments by learned economists who frequent this site!
Homer,
If the particpants in the IFI program are a random selection from all inmates, you have to include both those that stayed in and those that drop out, otherwise the results will always be biased in favour of the program. The key to thinking about this is that the control group also consists of those that would have completed the program if they had done it and those that would have dropped out. Within these sub-groups there is no evidence presented that the percentages of recidivism are different between participants and non-participants. Unless the sub-groups can be split for both program participants and control group, the overall average has to be used.
The case of students evaluating JQ’s course is not quite the same as evaluating the effectiveness of a program.
Mark,
If you have ever tried to do any research on prisoners then you would know that it is impossible to gain a random sample because of high risk prisoners and other things such as parole which was the main culprit in this study.
It would be similar in Australia.
Thus the reason for the ‘failure’ of IFI was the people who didn’t complete the course. A bit of a circular argument as it the main reason for the control group’s results. funny that.
In this case The IFI goal is to “re-educate” men over the course of a year, men who have spent most of their lives breaking the law, and we want to make them think and then live in a totally different way. IFI seeks to undo the habits of a lifetime. This I think you would agree is a long term goal.
Moreover what Professor Kleiman omitted in his Slate article and I alluded to previously is his support for education /literacy courses for prisoners.
He says “The first found that education course completion among federal prisoners was responsible for a 15.7% reduction in re-arrests, even after accounting for a number of other factors known to predict recidivism. The second found that Wisconsin inmates who complete a high school or adult basic education course are
20% less likely than the average offender to be re-incarcerated, again controlling forcharacteristics that predict recidivism.”
The only problem was these results were from prisoners which err COMPLETED the course not that started the course. OOPS.
( I actually support him in this.)
Sounds like we either completely give up on doing any research or like IFI (and unlike Professor Kleiman) insert the problems to engage in discussion.
Incidentally I have found the same problem cropping up in evaluating education programs in schools.
Brian, I did get involved in the McGovern discussion, but I’ve forgotten the details.
I do know we export the majority of our production of sugar, which isn’t on the list you cite. Beef, wool, wheat and sugar are our big production items, so if we export the majority of output in all of these, exports must be important.
Homer can you give a link for the Kleiman quote?
If you have ever tried to do any research on prisoners then you would know that it is impossible to gain a random sample
If so, you need non-experimental techniques because selection bias will be pervasive. It thus becomes even harder to work out what, if any, actual difference a program made.
Homer, there has been a solid 30+ years of highly technical work on just these techniques (James Heckman won a Nobel for his contribution to them). And it’s an area which is still highly contentious.
The bottom line here is that the fact that those who completed the course had better outcomes than those who didn’t really tells us nothing in itself about the course’s effectiveness. We have to very carefully work out whether the course contributed to the better outcomes, or whether it just inadvertently selected those who would have had better outcomes anyway. Its not always obvious which happened, especially when bad prospects tend to drop out at a different rate to good prospects.
Might I gently suggest that program evaluation is an area best left to specialists?
A correspondent to the AFR suggested that a good way to make politicians more responsive would be to have rolling by-elections, say 6 every month so that the whole Parliament faces the electorate over 2 years – you give them a year to settle in and then the random by-elections start. Just like in the real world, they would never know if they would have a job in a month’s time.
Regarding Steven’s earlier comment:” It’s not Australia’s own population that is the source of the overuse of the resource; we export most of our agricultural product, ie the transformeed water, soil etc”. Others have added the figure of 50% for exported product. So if agriculture accounts for for 80% of water use in the Murray-Darling basin, then 40% of the Murray-Darling water is used for trade. With hundreds of thousands of animals dying during transport, it makes me wonder how far reaching our water resource efficiency is going to have to become.