Utilising the Budget Standards of BSU - UNSW SPRC

return to menu


The long awaited Report of the BSU has now been released as a 600 page document. On the one hand it reports on far more than was intended by recommendation 116 of the JSC in November 1994 and on the other hand it does not attempt to directly answer the specific topics of rec 116 which related only to "a critical evaluation of the current Child Support formula percentages"

Possibly the futility of such an exercise was realised at an early stage in the evaluation and it was determined that the best path was to simply provide a wealth of data upon which others would be "enabled" to further analyse the overall situation. That is what is performed in this article.

There is considerable discussion in the report about the use of spreadsheets to present the data in order to allow customisation by spreadsheet analysis. The report then goes on to say, in speaking of the analysis of costs of children of different ages over different times, that it may be too complex to analyse and conclude such results as we are all very keen to know, ie the validity of the CSA formulae.

Unfortunately this assertion of difficulty may well be so (to use the Microsoft example) for those using post 1990 versions of Excel at which time (Excel 5) the application was basically made to look like "a word processor with cells" apparently in order to increase sales by enticing the non expert user. While a lot of the analytical tools which were dispensed with in the later versions can actually be found in hidden menus, there are 2 problems. Firstly even those skilled in Excel 4 find roadblocks where a critical feature was not hidden but deleted and secondly those that started using spreadsheets after 1990 had no idea that such tools and methods even existed, this situation being further frustrated by the fact that the excellent manual of version 4 which boasted more pages than the BSU Report was withdrawn from publication.

The author uses Excel 4 and is not constrained by this problem and as such asserts that complex analysis is indeed very possible as will be seen, simply by using macros and lookup tables - 2 of the casualties of Excel 5. In summary it could be said that the BSU data is readily usable for customisation as suggested in the final chapters of the Report.

How to use the Report

This site already details the possibilities to be had with the publication of the BSU results. The long term aim might be that seen by the JSC via Rec 116 but as Parliament has decided to make minimal changes to the CSActs BEFORE the BSU Report was fully digested, one would have to assume that it will be a long time before any intelligent changes are made to the CSA formulae.

However there are several other paths one may take immediately to obtain relief from a CS Assessment with the assistance (on all sides) of this Report.

The Avenues

  1. By application to the CSDO under Part 6A of CSAAct
  2. By application to the Family Court under Part 7 of CSAAct
  3. By application to the Family Court under Part VII of FLAct

It is not seen as desirable or appropriate to burden this article with legal jargon, but rather to simply state the "state of the law" in easily understood terms.

An administrative assessment by the CSA is nothing more than a starting point of what might be an acceptable determination of "proper" child maintenance in a particular case, but if it is not contested the assessment stands as a legal liability.

Once one decides to question an assessment however, and gets to "first base", then the inquiry which the FC or CSDO is required to make (even though the CSDO is recalcitrant in this respect) is one which examines the proper needs of the children and the ability of both parents to cater for these needs. Essentially parents on a high income are not expected to spend more than that actually required to support their children and parents on a low income are expected to do what they can together with the assistance (on all sides) of the DSS safety net.

In saying the CSDO is recalcitrant draws attention to the most important issue of all in regard to the BSU report which is that it is well settled by the court that, once the "departure" exercise is commenced, all reference to assumptions made via the CSA formulae is bypassed, except for the fact that to save parties the need to return year after year to the court the final result may be expressed as a "forced" change to one or more of the elements of the formulae.

Essentially a court considers evidence of the parties as to these costs but is also permitted to consider published data on the "cost of children" to "temper" any pleadings which might appear to be out of the ball park and offered purely on the fact that perjury is "allowable" in the FC. However as per Gyselman it now seems that "cost of parents" information is also allowable to temper parental claims.

The CSDO is meant to follow the same path but in its present quasi judicial capacity without internal or external appeal mechanisms a culture has developed which simply ignores matters settled by the court.

It is the intention of this article to get as quickly as possible to the "good news" as per such pleadings set out above under "the avenues" and to do so primarily with reference to actual cost of children research. However the subject 600 page report also details the other side of the coin required by the various acts which is the ability of the parents to provide such maintenance. This in effect introduces even more data into the mix of what we are trying to establish. For example the Report infers that an unemployed person (female) requires some $15,000 pa to survive but, as we know, such a person only gets about $9,000 pa from the DSS safety net (whatever its form).

At this site it has been decided to use the $9,000 figure as the amount needed by BOTH parents to "support themselves". Whilst it may be arguable that the new figure of $15,000 is more appropriate, a decision has been made to not introduce "further factors" by using such data. It is accepted that there are one hundred and one reasons (on both "sides") why the dole (as an example) does not cater for such costs but the author does not wish to presently include such for the reason that it may cloud the issues at hand. It may well be proper to include such data in future but the results hereunder are based on the figure of $9,006 (as it now stands) for the protected earnings or non liable income element of both parents.

So although the various models and modules do in fact contain many of the DSS and ATO algorithms applicable to family law situations (as well as a bit of "social engineering" of the family payment), they are not discussed at present.

The Harris Family

The family chosen for the first exercise is the Harris family used at the Child Support Bureau Service.

Or just look at the family itself and BACK to here. Please note that the CSA start date was chosen differently to that used here so the total liability is different. But before we get back to the example of the Harris family of 4 kids we need to establish the analysis tools and methods used to examine their "properness" under an administrative assessment.

In respect of "cost of children" information the BSU Report provides 2 levels of "costs" defined as "low cost" (LC) and "modest but adequate" (MBA). Within both levels there are 9 categories of expenditure :

  1. Housing
  2. Energy
  3. Food
  4. Clothing
  5. HGS
  6. Health
  7. Transport
  8. Leisure
  9. Pers Care

The computer algorithms set up by the author propose to deal with the BSU data in the following way.

First of all a "residence ratio" is established based on the court orders for contact or on that actually happening. The court defined term of "normal access(/contact)" stipulates 2 days contact per fortnite and 4 weeks holiday contact per year. This, when input to the author's software module, reveals a ratio of about 20% residence by the non residential parent. Other arrangements, when input to the module, will establish the appropriate ratio.

The "cost of children" to the parties is then established as follows :

The items identified as 2, 3, 5, 6, 7, 8 and 9 above are apportioned to the parties as to the residence ratio. It is suggested that all of these items provide "incremental liability" on the parties in proportion to their residence ratio. Item 4 is clothing and it is asserted in this analysis that the residential parent pays all of such budget.

The remaining item to be considered is that of housing and in this regard it is proposed to disregard, for the present, the wider issues as to property division and the effect that such may have had on child maintenance issues. In other words it is assumed that the court has not or will not "double dip" by awarding an additional "Dench" adjustment to the pool of assets of the parties.

It has been established in the family court that children should expect the same standard of accommodation with both parents. For this reason (and the reason that the children can not take their bedroom with them when on contact) both parents are provided with the same allocation under category 1. In effect the total figure will be higher than for the intact families in the Report and it is generally accepted that when parties separate the same income has to be stretched further.

The next decision is when to use the LC figures and when to use the MBA figures. An examination of the derivation of the figures on pp 593, 594 of the report indicates that the main difference between the 2 levels is childcare (in the HGS budget). The girl aged 3 does not need childcare in the LC because both parents are unemployed and have "time on their hands" as the Report points out, whereas both adults work in the MBA case, necessitating childcare.

Therefore the author has decided that the logical "toggle point" from LC to MBA is a point where a person's employment time does not enable them to provide their own childcare. The sole parent pension cutoff point has been selected as it allows a person to undertake some part time work without encroaching on childcare duties. It is also noted that the HGS budget applies to both parties as a working contact parent with only 4 weeks holidays a year is often faced with the need for full daytime childcare in the common scenario of orders at present which state "half of the school holidays".

It has to be said that it is refreshing to see the total transparency of the BSU approach in this Report as opposed to the clandestine approach of the AIFS (and the ensuing CSA formulae) where the question of childcare was possibly(?) covered by the formula percentages but was double dipped by the strange method of increasing the custodian's "exempted" income by about $30,000 in excess of that of the other party (and calling it "disregarded"). Although this differential is shortly to be decreased to $20,000 the author is of the opinion that it is misplaced logic to use such a device to account for a cost of children item which is simply one item on the list of several thousand which make up the total. However, as pointed out above, once a "departure" exercise gets to first base, all such parameters as disregarded income are in fact disregarded and the Coon and Cox "equitable" type algorithms employed in this exercise are substituted.

The Report cautions the user to observe the "bedroom rules" used in the housing cost Report and while it is possible with Excel 4 to devise an equation to replicate this requirement a simpler method has been devised which does not inquire into the gender of the children but which still would provide accurate results. The method uses a lookup table which provides one bedroom for up to 2 children, 2 bedrooms for 3 or 4 children and 3 bedrooms for 5 or more (as per the Report).

The main reason for the simplification is to do with the possibility that the CSA might eventually be legislated to use a formula based on the method herein and to ask the CSA programmers, presently up to their necks trying to remove the year two thousand virus they initially inserted, to then introduce gender considerations would seemingly see the CSA computer system crash completely. Also as suggested in the Report it is valid to use the figures for the 3 year old girl, for example, for a 3 year old boy and as seen below this simplification allows a method similar to the "regression" method on page 610 of the Report to be used.

In fact that explanation virtually completes the background and we can now go to the actual analysis. A lookup table is constructed for both the LC (page 593) and MBA (page 594) figures

Please take a look at the LC table (look21) and BACK to here

The yellow is the actual lookup table and the top line indicates the column numbers of the table. The left column is an explanation of the ages the table uses, eg the top line is for ages 0 to 4. The bottom line simply provides a zero result for a child reaching 18. A similar table exists for the MBA figures.

take another look and BACK to here

Then we have the summary for the children in the example

Please take a look and BACK to here

The yellow cell applies to the cost of Jane (apart from housing) to Mrs Harris and the equation in that cell is :


If that looks complex it simply says to go to U415 and find how old Jane is at the time of the "assessment" and having found she is 15 then go to the line of look21 with 15 in col 1 and multiply all in cols 4,5,7,8,9,10 and 11 by 0.7912 (the residence ratio in U415) and then add the lot of col 6 (clothing). Easy?

The green cell is the addition of similar amounts for the other kids plus 2 times $32 for the 2 bedrooms. In fact this seems to be a case where the bedroom rules require 3 bedrooms if the full gender analysis was to be conducted. However bearing in mind that the other party also gets the same allowance there would only be minimal error in using the suggested shortcut. So the cost to Mrs Harris is $282 per week.

In this example Mr Harris uses the MBA table and the cost to him is $150 per week.

So far so good, we know what the children cost each parent. The next step is to divide the cost equitably. In Coon and Cox Nicholson CJ simply used the ratio of the (before tax) wages of the parties to decide that the wife paid 66% and the husband 33%. It is considered that a fairer method is to first subtract a non liable income for each party (as suggested in Mee and Ferguson) and this amount has already been discussed above but left as $9,006. Also this amount is subtracted from the "after tax" income of both parties. However, like all the methods used herein, the author is open to any suggestions to improve the methodology - provided that the suggestion to not involve bias.

The author developed the algorithm =G46-(VLOOKUP(G46,look10,2) +VLOOKUP(G46,look11,2)*(G46-VLOOKUP(G46,look12,2))) -IF(G46<$S$42,0,$S$43*G46) in Excel 4 to determine the after tax income of the parties and this value is factored into the analysis. In the opinion of the author, to suggest (as is suggested) that the might of the "Patrick Stevedoring ATO (CSA) programming union" is unable to do same (while also devirusing their year two thousand virus which they themselves implanted, for their own reward) is "a nonsense" and not worthy of discussion in this article.

The comparison

The results are provided in the same format as for the Wright and Snag families at this site.

Please take a look and BACK to here

It is not the intention to delve into DSS type issues in this article though certain anomalies will be evident from these results. So to summarise the child maintenance situation Mrs H has a "cost of children" of $282 per week and Mr H $150, a total of $432 per week - all this being per the BSU figures as interpreted and analysed by this article.

The administrative assessment (Mr H earns $48,000 pa and Mrs H $18,000) requires Mr H to pay $255 per week to Mrs H as his supposedly correct contribution to her $282 cost, while she pays nothing towards his costs.

The "equitable solution" suggests Mr H pays $226 towards Mrs H's costs of $282 (80% of the total) and Mrs H pays $30 towards Mr H's costs of $150 (20% of the total), the "differential" result being that Mr H pays $197 per week. Please do not confuse the 20% with the 20% of contact time. The above 20% is purely associated with the relative abilities of the parties to support the cost of children of the other party.

It is hard to imagine that any person or authority would argue that Mr H's assessment is "proper" (via either the FLAct or CSAAct) so it is proposed to move straight along to the solution.

The solution

Even though one is required to completely disregard the formulae in assessing if an assessment is proper, once it is determined to be improper the remedy can then revert to a "manipulation" (in isolation or combination) of most of the paramaters of the formulae. The reason, as already stated, is the obvious need to avoid yearly court or CSDO intervention if the formula can simply be "reworked" to provide a proper ongoing solution.

So back to the Excel 4 analysis. The first solution which occurs is to reduce the percentages. Perhaps the most powerful analysis tool hidden (if not lost) since Excel 4 is Solver. Via Solver one can simply ask "if I pay $255 pw at 34%, what percentage is applicable to $197 pw?"

Please take a look and BACK to here

You will notice that the analysis has not only worked out the initial percentage as being 26.23% but has done a pro rata reduction of all the percentages, much in the same way as anticipated via s 125 (3)(b) of the CSAAct. Now it will be seen from the BSU data that apart from the housing element there are no great economies of size of family to be had from an increase in number of children. Therefore one would need to question the applicability of the CSA percentages (eg 3 children = 32%, 4 children = 34%) as a basis for a pro rata change.

However the Report also makes the comment that this situation will be influenced by the changes in the ages of the other children as each reaches 18, but concluding that such analysis would be very difficult. Not to be put off by such declarations of doom the author has simply written a macro which progresses though the "turning 18 type happening events" of the family under consideration. The method does a new assessment (in today's dollar) on the day after each event and plots the results on a graph.

Please take a look and BACK to here

In the graph the red line is the CSA admin assess which prescribes a total liability of $155,749. The green line is the pro rata reduction where the change in the assessment from 34% to 32% is mirrored by a change from 26.23% to 24.69% etc giving a total liability of $120,160.

Then by running the "doom" macro described above we are provided with the blue line which describes the actual "regression" as per the BSU data. The result is that the liability is reduced to $91,459, some $30,000 less than that suggested by a pro rata method based on 4 BSU children and a CSA regression.

Finally a strict linear reduction/regression is plotted in yellow which is remarkably similar to the blue figures.

The result is that the blue figure is 59% of the red figure for an administrative assessment. It is by no means suggested that the specific results from this exercise should be enacted into strict formulae as we presently need to address via the CSA. Rather it is suggested that the models developed should be used in ongoing research to "repair the formulae" as suggested by the JSC.

However, on the basis of the above results and conceding that even a "low cost kid" gets a share of 1 cent a week to the cost of a doormat, it would have to be suspected that the CSA formulae produce somewhat excessive liabilities and as such leave themselves open to all avenues of challenge.

It would seem that the intention was to provide formulae which would be equitable in the vast array of instances, without any need for "appeal". However it would also seem that these BSU figures simply confirm that the formulae produce a result which is wildly excessive or equally inadequate, thus tying up the resources of the court and CSDO - the very occurrence which was initially to be avoided by the introduction of the formulae into legislation.

In the opinion of the author, no repair is possible to the formulae whist the objects of the Acts talk to satisfying the needs of the children and the formulae talk to taking a percentage of income in apparent satisfaction. However in the meantime the various paths of the Acts allow individuals to plead their case using the above methods, BUT NOW with the assistance (on all sides) of the BSU figures.

return to menu