A reflection on expected experimental outcomes – Latent Semantic Analysis of interviews

When I first thought of the possibility of using Latent Semantic Analysis (LSA) to analyse my parent interviews exploring school choice my expectation was that the odds of success were slim.  I was familiar with the use of LSA in helping decipher the contextual meaning of words in ancient texts and indicate the likelihood a text was written by a particular person based on their known works.  LSA is one of the methods used to try and identify (speculate) who the ‘real’ Shakespeare was by comparing the works of Shakespeare with the writings of other contemporaries.  But could LSA be applied to interviews investigating economic decision making behaviour and provide meaningful insights? Would it be possible to identify key concepts influencing the decision making process based the common use of key words? Could LSA identify ‘latent authors’ representing distinct, heterogeneous, types of decision making within a society? Noting that the conventional economic wisdom is that society is comprised of a homogenous set of individuals (one type) applying the same decision processes subject to variability environmental conditions (such as wealth) and uncertainty.  To draw on quantum physics – there are no ‘flavours’ or handedness in standard economic theory.

Continue reading

Demographics of parents interviewed about choice of secondary school

The table below provides some detail as to the level diversity of the interviewed parents.  It is important to note that the focus of the study was to explore and understand the decision architecture of how parents choose a school for their children.  Consequently, a degree of priority was given to finding parents who had switched schools from public to private (independent/Catholic) or from private to public in the transition from primary to secondary schooling.

demographics

Non-Catholic private schools in Australia are called ‘independent schools’ because these schools (particularly the elite schools) govern, manage, resource and finance themselves independently of any broader religious or philanthropic affiliation.

In contrast, Catholic schools are managed, financed and resourced centrally in a manner broadly similar to government run public schools but with higher levels of community involvement and direct fee payment associated with the choice of school. While there has been a trend for Independent secondary schools to move from single sex to co-educational, Catholic secondary schools in the metropolitan cities by and large are single sex schools.

Continue reading

Latent Semantic Analysis

An everyday application of Latent Semantic Analysis (LSA) is the Google search engine where words that are semantically/contextually similar are also returned in the search query.  Type in “run” and the search will also pick up “ran”, “runs” and “running”. LSA allows natural language processing of vast collections of data, such as web pages, to provide information about how similar words are related to each other in (semantic) context by converting words into vectors (vectorial semantics) and applying singular value decomposition to the matrix.  In this way, the data itself is used to create a ‘latent semantic dictionary/thesaurus’ which reflects the context of the documents being analysed.

LSA captures the contextual relationships between text documents and word meanings.  Taking into account the context in which words are used is important for linguistic analysis.  The contextual meaning of words change over time and across social groups.  An example of the importance of context is how the meaning of ‘terrific’ changes over time.  Latent semantic analysis of documents from the second half of the 19th century would show ‘terrific’ as similar to ‘horror’. While documents from the second half of the 20th century would show ‘horror’ as now being the opposite of ‘terrific’.

Continue reading

Python Project – linguistic analysis of interviews

To help analyse my exploratory interviews into how parents choose schools for their children I’m teaching myself how to implement text processing in the Python programming language.  My objective is to apply Latent Semantic Analysis to the interviews to help understand how preferences are formed, relationships between these preferences and whether there is any linguistic grouping of preference strength indicated by their semantic distance to choice groups.

However, I need to undertake some text pre-processing for these analyses to work.  The pre-processing stages are:

  1. Remove punctuation
  2. Make all words lower case (so processed as same word irrespective of case)
  3. Remove stopwords (common words as “in”, “at”, “the” etc)
  4. Convert words to their root lemmas (eg. “run”, “ran”, “runs”, “running” to “run”)

Lemma processing is important for latent semantic analyses of relatively small text sizes of 10,000 to 100,000 words compared to the more usual many millions of words involved in web search algorithms.  For text sizes in the many millions of words Latent Semantic Analysis will do the equivalent of ‘lemmerizing’ as part of its ‘similarity’ processing.

I tried to use the ‘standard’ lemma dictionaries but found that I needed to categorise words more specifically such as grouping all the different sports together as ‘sport’, all the languages as ‘language’, all musical instruments as ‘music’ etc.   There are also words that are similar in writing but not in meaning that will be stemmed in a way that is inappropriate.  For example ‘certain’ & ‘certainly’, and ‘actual’ & ‘actually’ are not similar in the context of decision making regarding school choice interviews.  Even though linguistically they do have the same roots.  So in the end I created a dictionary that allowed me to categorise words in a way that was more meaningful and concise for the context of the analysis.

The best everyday example of how latent semantic analysis is used is Google.  Type in “run” and the search will also pick up “ran”, “runs” and “running”.

An example of my custom dictionary for lemmatising school choice text is provided here.

Factors influencing academic outcomes

Recently Gary Becker in the Becker-Posner blog opined that “I believe much of the blame rests with the fact that many children from minority families are raised with a single and not very educated parent, and that the quality of the schools attended by minority children is deficient.”  This is quite a pessimistic statement considering it has been 50 years since the publishing of Becker’s seminal work on the economics of human capital ‘Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education’ (1964).

Below is a high level summary of the factors effecting academic outcomes associated with school choice.

Factors influencing academic outcomes

Factors affecting academic outcomes and generational mobility are complex and highly endogenous. I will add a number of posts on this specific topic over time as I try to tease out all the factors and try to draw logical and consistent connections between these factors.

==========================================

Becker, G. S. (1964). Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education.

The Eusocial and the anti-Social

The eusocial and the anti-eusocial

I captured a photo of the eusocial Honey Bee Apis mellifera and parasitic Chequered Cuckoo Bee Thyreus caeruleopunctatus together on oregano flowers in my garden today.

The Chequered & Neon Cuckoo (Thyreus nitidulus, also common in my garden in summer) bees parasitize the nests of the solitary Blue Banded Bee Amegilla cingulate.

I wonder though whether the Chequered bee and the Neon bee are in fact morphs of the same species similar to polymorphism in the Monarch butterfly which has a white morph resulting from apostatic selection.  Both are basically the same, even the spotted pattern, except that one has white spots and the other bright neon blue spots.  The frequency of the Chequered bee and Neon bee seems to be inversely related, with temperature as the determinant.  Of the two, the Chequered bee seems to metabolically prefer hotter temperatures.  Using a refrigerator test, the Neon bee was clearly more responsive at lower temperatures (no bees were harmed in the testing).  Also this year was hotter earlier than last year resulting in a switch, at least in my garden, from Neon bees being more frequent last year to Chequered bees being more frequent this year.  If these two bees are in fact morphs of the same species, a reason for the phenotypic colour morphs could be due to Neon bee being regularly harassed by large black wasps with a similar neon blue glow.  This harassment seems to be of a mistaken sexual nature rather than predation but has an energy cost of continually buzzing the wings as a defence compared to the Chequered bees that aren’t harassed.  The black wasps are more common with hotter temperatures.

Neon Cuckoo Bee (s)

Thyreus nitidulus

Chequered Bee in flight (s)

Thyreus caeruleopunctatus

Blue Banded Bee (s)

Amegilla cingulate

The drunk neo-classical economist and the lamp post

Alan Greenspan’s recent article ‘Never saw it coming’ reminded me of the ‘drunk and the lamp post’ joke:

A drunk loses his keys and is looking for them under a lamp post. A policeman comes over and asks what he’s doing. “I’m looking for my keys” he says. “Where did you lose them?” the policeman asks.  “I lost them over there”. The policeman looks puzzled. “Then why aren’t you looking for them over here?” “Because the light is so much better here”.

For neo-classical economists these ‘lamp posts’ are mathematically elegant and tractable models, sometimes supported with econometrics, which lead to unambiguous conclusions. While these models are illuminating and logically consistent within themselves they are frequently ‘mugged by reality’, as Greenspan puts it, making these models a poor basis for forecasting.

This is an ‘observational bias’ where people focus their attention on areas leading to easily illuminated results.

An example in banking & risk management is the over reliance on risk models such as Value-At-Risk (VAR). Which, while mathematically complex, conveniently produce an ‘output’ which is easy to interpret and explain. This leads to an overconfidence in mathematical models measuring a bank’s risk exposure while the more dangerous risks, in hindsight after the GFC, are the organisational behavioural risks which tend to be largely ignored (in the shadows).  Including ‘model risk’ arising from cognitive biases affecting how models are constructed and interpreted rather than mathematical error per se.  Well known mega-losses where sophisticated risk models were mugged by behavioural reality are: Societe Generale’s rogue trader (USD$7billion), Morgan Stanley’s ‘hedging error’ (USD$8.7billion) and the bankruptcy of Lehman Brothers.  I worked for 2 of these 3 banks and turned down an offer from the third.

There is also an ‘observer bias’ where how we see things is not independent of our own condition, preferences and context.  Neo-classical economists tend to view all individuals in their models as being as rational and mathematical able as themselves even though in reality individuals tend to have a wide range of cognitive ability.

“We don’t see things as they are, we see them as we are.” Anaïs Nin

A rather brutal joke I often heard in banking, mostly from traders, was:

Q. What do you call an economist making a forecast?

A. Wrong.

The complexity of educational choices made by parents for their children

At a high level, choice decisions relate to trade-offs between consumption and savings, now and across subsequent time periods subject to constraints and uncertainty.  For parents, educational choices for their children are constrained by the parents’ income, time and regulations, and subject to high levels of uncertainty over very long time frames. Parental choice relating to investments in their children’s education only really occurs in the broad range of the socio-economic ‘middle class’. For the very wealthy, choice is the default of ‘only the best’ which requires little to no effort in decision making despite the cost of the education itself. Parents in low socio-economic conditions lack both the time and experience to research education options and the monetary resources to capture opportunities as they arise, leading to an acquiescence to the default choice of no action.

Rational choice theory suggests that parents are utility maximisers who make decisions from clear value preferences and can be relied upon to make decisions in the best interests of their children (Becker & Tomes 1976, 1979). Yet in deciding which school a child should attend, under rational choice conditions, a parent is required to make a series of complex intergenerational and intertemporal choices that would challenge seasoned economists. Educational choices are predominantly path dependent, subject to imperfect information and in most cases irreversible. Ordinary parents however, need to make these decisions with little training and with limited time to evaluate options. Instead, parents rely on a suite of behavioural heuristics in order to achieve a good outcome for their children.  Individual choice is also context dependent, subject to the experiences of parents, their expectations of the future, a duty to their children and emotional attachment.

How can a parent make optimal decisions in the face of so many possible choices and outcomes? Choices which are necessarily sequential and irreversible once made. To overcome the complexity of choice, humans have developed decision strategies which allow shortcuts to be taken to achieve a ‘good’ outcome in the face of incomplete information and limited time for evaluation. These heuristics, intuitive decision rules, allow mathematically hard problems to be solved under restrictive conditions where a good outcome is achieved at the expense of a perfect outcome. For a parent, a perfect outcome is only possible by chance and impossible by deliberate calculation.

While heuristics are ‘quick & dirty’ solutions, they draw on highly sophisticated underlying processes. Tversky & Kahneman (1983) testing the conjunction rule in likelihood rankings using the classic ‘Bill & Linda’ experiments showed that there was no difference between naïve and sophisticated participants. Experiments undertaken by Gigerenzer & Goldstein (1996) tested the effectiveness of fast and frugal decision heuristics, such as ‘take the best, ignore the rest’, against sophisticated statistical estimation strategies, such as Bayesian networks. Their research showed that fast and frugal heuristics did not fall too far behind a Bayesian network approach. More interestingly, as the quality of available information used for estimation decreased, heuristic strategies became more effective when compared with the more sophisticated strategies.

The complexity of the decision architecture associated with making choices, combining both rational choice and behavioural components, is illustrated in the ‘Choice Process’ diagram below:

The Choice Process

————————————————

Becker, G. S., & Tomes, N. 1976. Child Endowments and the Quantity and Quality of Children. The Journal of Political Economy, 84(4), S143-S162.

Becker, GS & Tomes, N 1979, ‘An equilibrium theory of the distribution of income and intergenerational mobility’, The Journal of Political Economy, 1153-1189.

Tversky, A & Kahneman, D 1983, ‘Extensional versus intuitive reasoning: The conjunction fallacy in probability judgment’, Psychological review, 90(4), 293.

Gigerenzer, G & Goldstein, DG 1996, ‘Reasoning the fast and frugal way: models of bounded rationality’, Psychological review, 103(4), 650.

McFadden, D 2001 ‘Economic Choices’, The American Economic Review 91(3): 351-378.

Heritability of academic outcomes with a surprising difference between girls and boys

The heritability of cognitive ability influencing school academic outcomes is well established and at a general population level Nature vs. Nurture influences are roughly split as 50% genetic (parents), 30% shared environment (family and school) and a remaining 20% unshared (uniqueness).

A recently published UK study of 11,000 twins  (Shakeshaft et al. 2013) produced some interesting segmentation of heritability across subject areas: English 52%, mathematics 55%, science 58%, humanities 42%, with the impact of shared environment being 36%, leaving uniqueness between 4-12%.  Overall GCE scores gave 53% heritability, 30% shared and 17% uniqueness.

The more fascinating result that came out of the study however, is that heritabilities are somewhat greater for boys (57%) than for girls (47%) and that shared environmental influences are greater for girls than for boys.  While this result is very interesting, the authors “prefer merely to note these significant sex differences in our sample and to defer speculation about their origins until these results are replicated, for reasons discussed later.”

Difference in heritability could be explained if boys had a greater variance in cognitive ability compared to girls due to sex differences in the heritability of autism and dyslexia etc., however the statistical variance for boys and girls are similar for this study. While the mean academic scores for girls are appreciably higher than for the boys in the study.

This would suggest that the Y chromosome somehow amplifies genes associated with cognitive abilities while at the same time impeding genes associated with social ability, leading to lower shared environment effects.  Social behaviour genes however, could give rise to stronger shared environment effects while at the same time generating significant group learning and problem solving benefits.  Leading to the conclusion that within the larger set of hereditary influences on cognitive ability there is a sub-set of heredity influencing aspects of sociality associated with the development of cognitive ability through group learning.  This hypothesis would be consistent with observations of differences in learning styles between girls and boys.

It is important to note that there is a fair amount of indirect heritability occurring through the ‘shared environment’.  A large part of the family influences can be attributed to the cognitive ability of the parents affecting their socio-economic status and thereby their ability to invest in their children’s education.

The significance of this indirect heritability can be seen in circumstances where children are adopted into a family with biological offspring.  In a study of African American children adopted into Caucasian American families by Scarr & Weinberg (1976) it was found that ‘Black’ children from parents of average IQ and low socio-economic status adopted into ‘White’ families of high IQ and high socio-economic status performed better than the average for ‘White’ children.  Biological children of the ‘White’ parents scored even higher on the tests.

Adoptive twin studies have shown that environmental & social impacts are greatest when adoption occurs when very young and where the children come from very low socio-economic backgrounds (Plug & Vijverberg 2003). This underscores the importance of government funded early childhood education in preference to school education when trying remediate socio-economic impacts on developing a child’s cognitive ability.

In the UK study, socio-economic background was not analysed so we don’t know whether there was any heterogeneity based on the ability of parents to invest in there children education.  It would have been interesting to see whether there was any reversal of heritability of cognitive ability for the lowest socio-economic groups  indicated by other studies (Turkheimer et al. 2003).

————————————–

Shakeshaft, N. G., Trzaskowski, M., McMillan, A., Rimfeld, K., Krapohl, E., Haworth, C. M., … & Plomin, R. (2013). Strong genetic influence on a UK nationwide test of educational achievement at the end of compulsory education at age 16. PLoS One, 8(12), e80341.

Scarr, S., & Weinberg, R. A. (1976). IQ test performance of Black children adopted by White families. American Psychologist, 31(10), 726.

Plug, E & Vijverberg, W 2003, ‘Schooling, family background, and adoption: Is it nature or is it nurture?’, Journal of Political Economy, 111(3), 611-641.

Turkheimer, E., Haley, A., Waldron, M., D’Onofrio, B., & Gottesman, I. I. (2003). Socioeconomic status modifies heritability of IQ in young children. Psychological science, 14(6), 623-628.

Isaac Asimov quote on scientific discoveries

“The most exciting phrase to hear in science, the one that heralds the new discoveries, is not ‘Eureka’ but ‘That’s funny…'” – Isaac Asimov

like, why is fundamental discounting behaviour (whether Rational or Behavioural) common to most multi-period investment & consumption decisions missing from decisions parents make about investments in their children’s education?  The ‘wait a sec, where is it?’ moment.

Intergenerational discounting & parental investment in education: Evolutionary basis for a zero discount rate heuristic

I will be presenting at the ‘Cooperation and Conflict in the Family’ conference – UNSW in Sydney, Australia from February 2-5 2014. This is a multidisciplinary conference drawing together evolutionary economics, sociology & anthropology.

ABSTRACT This paper discusses the absence of intertemporal discounting in human parent decision making behaviour associated with choices and investments in their children’s education. This behaviour is inconsistent with standard rational choice theory where parents should maximise the present value of the utility of their consumption choices over time. Nor is it consistent with behavioural economics’ expectation that individuals discount consumption choices across time periods hyperbolically. Parents applying a zero discount rate to the expected future returns from investments in their children’s education is however consistent with evolutionary theory. We propose that the proximate cause of this behaviour is a meta-heuristic, intergenerational temporal empathy, which is constructed from a core set of cognitive biases and heuristics so as to cancel out hyperbolic discounting behaviour normally associated with non-offspring investment related consumption across time periods. The ultimate cause of this meta-heuristic is to eliminate the propensity of hyperbolic discounting behaviour to under-invest in offspring development, thereby ensuring that inclusive fitness is maximised. Intergenerational temporal empathy also has characteristics of a positive feedback mechanism ensuring that beneficial educational strategies are propagated forward and may help explain divergent outcomes for low socio-economic groups where poor educational investment decisions tend to be reinforced across generations.

Social discount rate & very long term investment horizons

My interest in how discount rates are applied to investment decisions comes from my corporate background building and evaluating many business cases for large, long term projects where parameters, including the discount rate, are prone too often to gaming behaviour.  Particularly for government and semi-government organisations which don’t have the advantage of a straight forward Weighted Average Cost of Capital (WACC) calculation.

The recent paper by Gowdy et al. (2013) ‘The evolution of hyperbolic discounting: Implications for truly social valuation of the future’ made me think more about the vexed question of how the social discount rate should be calculated.  With exponential discounting, which is typically used in corporate finance, the present value of returns in future years approach zero fairly quickly.  This is problematic for social issues such as climate change, as indicated in ‘The economics of climate change’ (Stern 2007), the motivation for incurring remediation costs now is the expected positive social returns far out into the future.

Gowdy et al.’s view is that hyperbolic discounting should be used because it at least leaves a positive residual present value for long dated social returns in the future.  A hyperbolic discount rate approaches a value over time which is materially greater than zero compared to the conventional exponential discount rate that approaches zero. However, hyperbolic discounting is a behavioural heuristic which varies greatly across sections of the community.  Research conducted by Harrison et al. (2002)  ‘Estimating individual discount rates in Denmark: A field experiment’  indicated that hyperbolic discount rate between socio-economic groups can differ by around 10%, 32.9% for the poor verse 22.5% for the rich.

My view was why not just have a zero discount rate?

Interestingly in the paper ‘Are we consuming too much’, (Arrow et al 2004) the authors noted that Ramsey (1928) and Solow (1974) were of the view the social discount rate should be zero. A zero discount rate was rejected because it implies a savings rate of around 67% which economists including Arrow thought was not realistic. However, if you assume the cost of bringing up kids forms part of the investment/savings requirement, this would account for 50% outright of the 67% (leaving 17%) savings expected when the discount rate is zero. ‘Cost of raising a child in Australia’ summary data : 50% of income ~ 3 kids, middle income, after tax of 30%.

Treating the cost of raising children as ‘savings’ then makes applying a zero social discount rate to long term, inter-generational, decisions reasonable.  The difference between an optimal social consumption discount rate and zero being the ‘positive externality (outside the marketplace) from the welfare of future generations’ (Arrow et al. 2004).

———————————————-

Arrow K, et al. 2004 ‘Are we consuming too much?’ The Journal of Economic Perspectives, 18(3), 147-172.

Gowdy, J., Rosser, J. B., & Roy, L. (2012). The evolution of hyperbolic discounting: Implications for truly social valuation of the future. Journal of Economic Behavior & Organization.

Harrison, G. W., Lau, M. I., & Williams, M. B. (2002). Estimating individual discount rates in Denmark: A field experiment. The American Economic Review, 92(5), 1606-1617.

Ramsey FP, 1928 ‘A mathematical theory of saving’, The Economic Journal, 38(152), 543-559.

Solow RM 1974 ‘The economics of resources or the resources of economics’, The American Economic Review, 64(2), 1-14.

Stern, N. (2008). The economics of climate change. The American Economic Review, 98(2), 1-37.

Academic rank order more important than academic ability in determining educational outcomes

A recent discussion paper by Richard Murphy and Felix Weinhardt at the London School of Economics, summarized in the article ‘Top of the Class’, suggests that a student’s academic rank in a school relative to other students strongly influences “non-cognitive skills such as confidence, perseverance and resilience” which in turn have a big impact on future academic outcomes.  This conclusion is based on a survey of some 15,000 UK students and matched against student test scores.  The authors found that rank order in primary school had a material effect on academic outcomes at secondary school.

Essentially, if there are 2 students of the same academic ability at primary school but one is ranked in the top 1/4 of an average school and the other ranked in the bottom 1/4 of an elite school, when the students get to high school the student with the high rank order in primary school will achieve materially higher test scores than the other.  This goes against the accepted wisdom of the importance of the student peer effect where is generally held that it is better to be in a school amongst high achievers than at a school with not so high achievers.

They suggest that the mechanism by which this divergent outcome occurs is that by being in top of the school cohort the student becomes more confident and thereby enjoy learning more, consequently leading them to spend more time improving their skills.  What is particularly interesting is that this rank order effect is more pronounced, four times more, for boys than for girls.

Personally this confirms my anecdotal observations growing up in country NSW. I could see that we always had our above average share of great sports people. I put this down to confidence through achieving and the mind set associated with a habit of winning from a young age. A benefit of being a part of many small population groups, thereby giving more of a chance to be a ‘winner’. Logically this effect had to be strong to overcome the benefits big cities like Sydney have in their advantage of large numbers generating, statistically, more genetically exceptional sports people.

School choice and Equity: OECD Literature Review

Pauline Musset’s OECD Education Working Paper ‘School choice and equity, current policies in OECD countries and literature review’ is a great review of research exploring motivations and educational outcomes of school choice policies, with a particular focus on evidence based research.  This working paper is one of the most lucid and succinct reviews of the very contentious area of education policy and its ability to effect educational outcomes.

Some of the topics covered are:

  • self-segregation based on ability, ethnicity or socioeconomic background
  • diversity of schooling
  • performance differences across school types and countries
  • the use of vouchers and the effectiveness of different types of voucher systems
  • how school choice can exacerbate social inequity

I found Pauline Musset’s classifications of the ideological motivations for increasing school choice opportunities for parents particularly useful.  She classifies these motivations into 3 groups:

  • introduction of market mechanisms in education to remedy inefficiencies;
  • individualist-libertarian claims of a parental right for choice in education;
  • school choice as a way of making education systems more equitable.

=============================

Musset, P. (2012), “School Choice and Equity: Current Policies in OECD Countries and a Literature Review”, OECD Education Working Papers, No. 66, OECD Publishing.

A discrete choice experiment : school choice

As part of my PhD research project “Modeling the hardest decision parents will make: School Choice” I will conducting a Discrete Choice Experiment (DCE) to understand how Australian parents apply behavioural decision rules in choosing schools for their children, and how choice sets are evaluated to find the ‘best choice’.

This research will help understand why Australian parents are spending an extra $5.4billion more per year in private school fees when socio-economic sorting across public schools would lead to the same academic outcomes.

We also hope to gain insight in to why some recently opened private schools have failed, leading to large losses, despite the strong willingness of parents’ to pay for private schooling.

This project will also look at the influence of a parent’s socio-economic background on the type and strength of preferences. Our results will have important social equity implications for understanding how wealth, occupation and prior education affect a parent’s choice of school. This will be the 1st time a discrete choice experiment has been applied to school choice even though it is an experimental approach applied extensively in other quasi-market public good areas of health, the environment & transport infrastructure. Previously DCE hasn’t been applied to education due to the inherently endogenous behaviour of educational choice attributes, which this project resolves.

If parents are not making classical rational choice decisions, this research will have important implications for economics & education policy that have not previously been identified.

Why do Australian parents pay $5.4billion more for education than they need to?

Why are Australian parents spending an extra $5.4billion per year in school fees when socio-economic sorting across public schools would lead to the same academic outcomes?

In Australia nearly a 1/3 of children attend non-government schools, however, analysis by the Australian Council for Educational Research shows that the type of school a child attends has little impact on academic outcomes.  School academic outcomes are primarily driven by the socio-economic factors of the parents and the student peer group of the school.  ACER’s analysis of the 2009 PISA results (Thomson et al. 2010) shows that the majority, greater than 90%, of a student’s academic results are the result of either their family background or student peer effects at a particular school. These results are consistent with the recent Grattan report ‘The myth of markets in school education’ that the choice of school type had little to no influence on academic outcomes, other than the student peer effects associated with the socio-economic composition of the school student body.

If we apply the classic community optimization problem  (Tiebout 1956), in a solely public school system, families in their attempt to optimize both wealth and academic outcomes will sort themselves spatially & demographically in a way that leads to optimal choice outcomes.  In this way the benefits of a child’s socio-economic background will be matched with a student peer group of a similar socio-economic background in order achieve optimal academic outcomes for families.  Below is a conceptual diagram of the sorting.

Tiebout like sorting

While the educational outcome maybe optimal sorting, there is no net gain for the economy as what is gained at the top is given back at the bottom.  This is consistent with econometric studies in the USA (Ladd 2002). What is optimised is the wealth vs. education trade-offs for individual families.

We know that this type of Tiebout-like sorting occurs naturally within an economy from econometric studies showing a correlation between residential house prices and school quality (Black 1999).  In Canberra, Australia, a study done by Davidoff & Leigh (2008) indicated that for every 5% increase in school average academic achievement prices of houses nearby increased by 3.5%.

Tiebout like sorting - private schools

Approximately 20% of school students attend Catholic schools in Australia and another 14% attend Independent schools.  If academic outcomes are principally a function of socio-economic factors and not school type, why are Australian parents paying an extra $5.4billion per year (2009 figures) to choose a non-government school?

Is this because school choice is driven primarily by signalling of student peer group socio-economic ‘quality’ and private schools provide better signalling of this ‘quality’?  Or is school choice is driven primarily by ‘consumption’ of non-academic preferences such as teacher leadership, culture, & personal development?

I hope to answer these questions by conducting an online discrete choice experiment into school choice as part of my PhD research.

The basis for my $5.4billion calculation is:

From waterfall Graph 5 of the Gonski report, in 2009 student school attendance by type for Australia was: Gov’t 2.3million, Catholic 0.7million, Independent 0.5million. Private/parental costs per student per type are: Gov’t $0.4k, Catholic $2.6k, Independent $8.2k Deducting the Gov’t baseline parent cost of $0.4k, $5.4billion = (700k catholic students x $2.2k) + (500k Independent students x $7.8k). Note also that these school cost figures are recurrent costs/income and for both primary & secondary schools. Capital grants are excluded, mainly because capital costs have a number of sources other than parents.

It is interesting to note that the total recurrent cost of teaching a gov’t school student is $11.1k per year (2009) and that Catholic schools have a lower cost of delivery at $10k per student for what you could say is a better quality outcome.

————————————————————–

Black, S 1999 ‘Do better schools matter? Parental valuation of elementary education’, Quarterly Journal of Economics 114 (2), 577–599.

Davidoff, IAN & Leigh, A 2008, ‘How Much do Public Schools Really Cost? Estimating the Relationship between House Prices and School Quality’, Economic Record 84(265): 193-206.

Ladd, HF 2002, ‘School Vouchers: A Critical View’, Journal of Economic Perspectives 16(4): 3-24.

Thomson, S, De Bortoli, L, Nicholas, M, Hillman, K & Buckley, S 2010 ‘Highlights from the full Australian Report: challenges for Australian education: results from PISA 2009’, Australian Council for Educational Research.

Tiebout, CM 1956, ‘A pure theory of local expenditures’, The journal of political economy, 64(5), 416-424.

Determinants of parental school choice: A Qualitative Study

I’m currently undertaking a qualitative study into the determinants of parental school choice.  The study comprises 22 semi-structured exploratory interviews of Australian parents,  principally from Melbourne with some from regional Victoria.  Parents come from diverse backgrounds of educational history, educational choice of school type, and cultural.

Following from George Shackle that for choice there need to be alternatives, the socio-economic backgrounds of the parents interviewed are broadly middle socio-economic, from low to high middle class.  For the very wealthy there is no alternative to the ‘best’ and for the low socio-economic parents income & behavioural constraints mean that there are no alternatives to their default choice.

Australia is a particularly interesting country to investigate the decision architecture of how parents choose a school for their children due to the absence of strong racial (USA) or social-class preferences (UK).  Race and social-class preferences are present in Australia but are not strong enough to completely outweigh other preferences that it is difficult to differentiate preferences associate with teacher quality, student personal development, discipline and community.  Race in the USA and social-class in the UK have become dominant proxies for these more differentiated preferences leading to simple binary choice decision making.

Having a range of differentiated preference attributes allow a deeper investigation into how social & family human capital investment preferences are traded-off between each other and the decision strategies that are being used.  When there a number of competing preferences which are quite varied, it can be very interesting to find some behaviours that should be present but is consistently missing from the interviews.  When choices are binary by proxy it is difficult to find decision structure and gaps.

Parental choice is extraordinarily complex being both inter-generational and inter-temporal in nature. It is subject to parental income, time and regulatory constraints. Choice is subject to high levels of uncertainty over very long time frames. Choices are path dependent, in most case irreversible and subject to imperfect information. Individual choice is also very context dependent, subject to the experiences of parents, their expectations of the future, a duty to their children and emotional attachment.

Where there is a common curriculum, choice is not about the quality of ‘education as knowledge’ per se. At a very high level, education choice is about quality of instruction, teacher quality or program choice, or the quality of the learning environment effected by student peers and culture. Individual preferences of parents are reflected in the type of school they choose for their child to attend. Choice may be a default choice based on constraints, or a choice exercise within a type of schools such as different public schools, or across school types.

The main education choice that this study focuses on is choice of secondary school by parents for their children.  Choices in Australia are: independent schools (generally Protestant religion aligned), Catholic schools, government public schools (entry determined by residential boundary), and government selective public schools (academic, music, sport etc.).  The state of Victoria also has an accelerated learning program in some publics schools which are not constrained by location of residence by require passing an entrance exam.

A comprehensive review of the Australian school sector can be found in the Gonski Review.

It is important to note that unlike the USA, funding of schools in Australia is from broad based taxes (universal vouchers approach) and not aligned with local property taxes.

%d bloggers like this: