Mythbusters - Does the Government waste money in June?

Author: 
John Sheridan - CIO & CISO

Have you ever rushed at the last minute to buy a Christmas present? I have and I bet I’m not the only one. Open data from our friends at the Australian Bureau of Statistics demonstrates this clearly:

% of Retail Sales by Month Australia

This isn’t surprising because we all know that a Christmas present delayed until the Boxing Day sales somehow isn’t the same. Of course, if you’re super-organised, buying in the Boxing Day sales for the next Christmas Day is a clever thing to do. It turns out that timing is pretty important in buying Christmas presents. But timing doesn’t give you much information about the quality of the present.  

My thesis is that this is true of government spending too. Contrary to the analysis one sees repeated occasionally, I contend that spending in June cannot be assumed to be wasteful when considered in context.

What is that context? Firstly, the Government has an annual budget cycle, based on a July-June financial year. With some exceptions, funds allocated to a FY should be spent in that FY. While not a perfect proxy, contract value approximates expenditure, up to a point. (For example, such spending doesn’t include APS salaries.) So, you’d expect to see at least some contracts begin and end in the same FY. And, indeed we do see that, to a certain extent. On average, over a seven-year period, 22% of contracts by value and 75% by volume end in the FY they began. As more expensive contracts can safely be assumed to last longer, this isn’t all that surprising. 

As an aside, my analysis is based on the historical contracts data set available on data.gov.au, which covers 1 July 2007 through 31 December 2014. I only considered contracts that started from 1 January 2008 through 31 December 2014. I did not use the records from 1 July – 31 Dec 2007 to avoid a bias so the data is in calendar years not FYs. Data from January – June 2015 wasn’t included as we haven’t finished cleansing it yet. So, the total contract value is $266 billion across 506,838 contracts. This data comes from AusTender on which all contracts at or above $10,000 are required to be recorded.

% of Contracts that end in the FY in which they start by month CY 08 - CY 14

 

  January February March April May June
Number 75% 77% 75% 72% 64% 40%
Value 13% 24% 19% 23% 21% 12%
  July August September October November December
Number 87% 86% 86% 84% 84% 80%
Value 23% 21% 25% 31% 34% 19%

So, we can see that when all contracts are considered, not just those with a very short duration (<31 days), June is actually not a peak period in which contracts start. 

What about when we consider all contracts, not just those that end in the FY in which they began? Here, the situation is even clearer. Most contracts start in July:

Contract Value by Start Month CY 08 - CY 14

Since budget funding, while often announced in May, regularly starts in the following July, this is to be expected. 

None of this is suggesting a June peak. How would one be achieved? Spending money quickly necessitates a quick procurement method. Generally, this requires a limited tender (or direct sourcing). Limited tenders are allowed for contract values below $80,000 and, in some circumstances, for larger contracts. About 40% of contracts above $80,000 by volume and about 58% below $80,000 by volume (56% by value) are conducted by limited tender. 

Purchasing off an established panel can also be quick but, historically, less than 6% of contracts by value go through panels so I’m going to ignore that. Open tenders need to be in the market for at least 25 days normally so June expenditure would need to be planned some time in advance.

Consequently, if there were a June splurge, we’d expect to see a peak in limited tenders less than $80,000 in June, but we don’t:

Even if we ignore the $80,000 limit, we don’t see such a splurge:

% of Limited Tenders by value by Month CY 08 - CY 14

Having looked at the context, let’s zoom into the data most recently discussed publicly – the value of contracts reported with a June start date with less than 31 days duration. Rather than just considering one year, I have used the average over seven years again. I have compared it with the total value of contracts reported in June and considered it as a percentage of both the monthly contract value and the annual contract value.

Average Monthly Contract Value based on Contract Start Date CY 08 - CY 14

 

  January February March April May June
Yes
($AUD Billion)
$0.05 $0.07 $0.08 $0.08 $0.13 $0.47
No
($AUD Billion)
$2.66 $1.88 $2.42 $1.91 $2.06 $3.15
% < 31 Days 0.27% 0.53% 0.43% 0.55% 0.85% 1.84%
%< 31 of Annual $ 0.02% 0.03% 0.03% 0.03% 0.05% 0.18%

 

  July August September October November December
Yes
($AUD Billion)
$0.10 $0.07 $0.08 $0.07 $0.07 $0.07
No
($AUD Billion)
$9.40 $4.24 $2.48 $1.76 $2.05 $2.62
% < 31 Days 0.15% 0.23% 0.44% 0.55% 0.47% 0.38%
%< 31 of Annual $ 0.04% 0.03% 0.03% 0.03% 0.03% 0.03%

I agree that there is a peak of short term spending in June. But it’s a pretty small peak. The June figure (that is the value of June contracts of duration less than 31 days) amounts to 4.5 times an average day’s value over the 2556 days in the period reviewed. That means this particular value in the month of June is 15% of an average 30-day period. The amount in question is 0.18% of the average annual value.

Of course, of that particular spending type (contracts of less than 31 days duration), relatively small that it is, the value of June contracts is about four times the average month. Like the Christmas presents purchased in the week before Christmas though, this says nothing about the quality of the spend. 

Delegates are still required to vouch that any expenditure is value for money. Given the annual budget cycle, monitoring spending closely, maintaining a reserve while it is needed, and managing risk are all valid and prudent budgeting measures. It’s hardly surprising then that their spending plans encompass such an outcome, particularly one that covers only 0.18% of the average annual value. 

So what do you think – is this myth busted? Your comments are welcome.

Regards
John

Comments (9)

I have an idea for government spending. At the end of each financial year when we as individuals submit our tax return, we decide where we would like our taxes spent. For example I might choose... 20% to the aged pension, 5% for disability care, 1% for unemployment benefits, 15% to the health system, 9% for education, 30% for infrastructure and10% for defence. The government gets a default portion of 10% to spend as they see fit. Just a thought...

Hi anonymous,

Participatory budgeting, which is very like your idea, is reasonably well developed in local government around the world (http://www.participatorybudgeting.org). It's been done at this level in Australia. There is even software available to do it (http://budgetallocator.com for example). I haven't seen any evidence of it being done at the national level anywhere though. 

Regards

John

Thanks John! Not such a novel idea after all :)

that's a limited take on government spending. if you look at the financial reporting by all Commonwealth agencies on a monthly basis, June is far and above all the other months. you limited analysis only looks at a narrow spending criteria.

so congratulations for using the data you want to find the outcome you desired - true evidence based work

Agreed asaa.
I have personally observed the splurge of funds in the last weeks of the financial year for both Departmental and Administered funds.

The best way to mange this would be to offer annual bonuses to Departmental Secretaries based on the percentage they brought their actuals in below budget for the year for both categories.

A budget is the maximum funding that may be required for the year including contingency. The goal should always be to respect the taxpayers funds and minimise expenditure.
To 'achieve' budget is a fail.

Success should be measured by delivering the required outcomes for the lowest cost. Finance and Treasury should also be more co-operative in rolling over funding to the subsequent year to discourage the end of year splurge.

asaa and Geoff - please don't shoot the messenger. John has clearly stated his parameters whereas asaa has given no reference for the scope of "financial reporting". I have also seen an apparent increase in discretionary spending in June but would contend that this is largely driven by the inability to carry any savings over FY's. I would also concur with John that any increase in this period is minor compared to total expenditure across the FY and your preferred interpretation is cherry picking of the type used to create headlines. As for bonuses being given to Secretaries based on how cheaply they deliver services - why don't we outsource the entire PS to an administrative services provider based offshore and just keep the Secretaries employed here. Imagine the savings! This kind of thinking fails to consider where the majority of PS funding gets spent - here in Australia, on people (staff), goods and services. Talk about a narrow spending criteria.

Some great discussions here and very contemporary, especially as the topic has been recently gained some media attention.

http://www.theage.com.au/national/public-service/everyone-except-the-gov...

Good to see some analysis of tenders data. I know how involved it is and the multitude of suppositions & assumptions that can come into play when calculating anything!

I'm providing a way to filter Commonwealth grants & tenders data at OpenAus.

The top-level functionality allows you to total both grants and tenders by postcode, key word, supplier/recipient name, council area and federal electorate. It's a complex beast and still evolving but I don't think there is anything else like it.

If you want to delve deeper into the tenders data visit the Tenders site or if you want to look at spending across the budget look at BudgetAus

This commentary does not deal directly with the pattern of expenditure at year the end of the financial year. This is largely because it attempts to address the matter through analysis of data from AusTender, a now antiquated system that was created and designed primarily to support accountability in sourcing.

The focus in this discussion should be spend analysis rather than the pattern of contracting. Spend analysis is the fundamental building block for category management and strategic sourcing. It focuses on opportunities to improve compliance, reduce costs and improve efficiency. The core data needed to support relates to accounts payable, vendor records and product and service classifications. In the Australian public sector some organisations may have internal systems that assist them in developing the requisite capabilities but all have the opportunity to buy in expert support from third party vendors.

Readers should ask themselves how it could be that commentary and discussion of this kind can proceed in the Commonwealth in the 21st century without any reference to spend analysis. Surely this is an indication of disengagement from the mainstream of professional procurement practice and thinking in both the private and public sectors!

Add new comment

Commenting

  • Allowed HTML tags: <a> <strong> <em>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

Last updated: 04 September 2015