Perception Versus Reality
The Actuarial Pulse is an anonymous, web-based survey of Institute members, run on a monthly basis, giving members an opportunity to express their opinions on a mixture of serious and not-so-serious issues.
What would you like to know? If you have a question you would like to put to the membership, email the Chief Editor.
REPORT GENERATED ON 13 MARCH 2014, 284 RESPONSES.
Interestingly, over 30 additional people opened up the survey, entered their age and gender, then felt that estimations were too much like hard work. Or perhaps they were unsure of whether their Fellowship would be taken away from them if they were found wanting, much like volunteering to trial the Part III examinations. Whatever the reason, they missed out on the fun.
As the Pulse survey began, ‘actuaries are known to be reasonable at estimating reality’. This Pulse was designed to gauge whether this is actually true, both in areas inside and outside of the profession.
The results are quite surprising, particularly our lack of understanding of expenditure on aged care and the aged pension. But before we delve into the results, indulge me while I provide some background to this Pulse. Recently I was reading about the concept of “wisdom of the crowd”, whereby the collective opinion of a group of individuals may provide greater insight than a single expert.
The classic example often cited was from 100+ years ago at a county fair in Plymouth. Similar to today’s ‘lolly guessing competition’, 800 people estimated the weight of an ox. The median guess was within 1% of the true weight of the ox.
Many studies have shown that a large group’s aggregated answers for quantity estimation, general world knowledge and spatial reasoning has generally been found to be as good as the individual. If you are interested in delving into this further, google James Surowiecki and his book on The Wisdom of Crowds.
Understandably, the studies have shown that crowds tend to make the best decisions when they are made up of diverse opinions and ideologies. Unfortunately, this survey picks solely on those with an actuarial background. Are we a diverse enough group to provide reasonable estimates?
SECTION 1 – ACTUARIAL RELATED INFORMATION
The first section of questions was more ‘actuarial’, so hopefully the respondents had some basis for their answers. All of the figures from this section were sourced from the ANPHA website (Australian National Preventive Health Agency) from their State of Preventive Health 2013 report.
Q1: WHAT PERCENTAGE OF AUSTRALIANS DO YOU THINK DRINK MORE THAN TWO STANDARD DRINKS PER DAY?
This question was broken up to receive responses separately for Australian males and Australian females. On average, 19.5% of adults consume more than two standard drinks per day, with males at 29% and females considerably lower at 10%.
And how did the actuarial profession stack up at estimating this? Well, the median estimate was higher for both males and females, but it was considerably higher for females. Our median estimate was 33% for males and 24% for females. Perhaps we associate ourselves with people who enjoy an extra glass or two of the amber ale.
In an attempt to show the diversity in responses, compared with the median estimate and the actual response I pulled together the following graph, split by the answers to Australian males on the top (in blue) and Australian females on the bottom (in yellow).
On the horizontal axis, I have grouped the responses into 5% bands and highlighted in green where the actual figure lies and highlighted in red where the median estimate lies.
Interestingly our female respondents had a median estimate of 40% of Australian males imbibing more than two standard drinks a day! However, when it came to Australian females, their estimates were slightly under the estimates of their male actuarial counterparts.
Q2: WHAT PERCENTAGE OF AUSTRALIANS DO YOU THINK SMOKE ON A DAILY BASIS?
Similarly to the last question, one of the next biggest preventable health issues was also examined separately for Australian males and Australian females. On average, 16.1% of adults smoke on a daily basis, with males at 18% and females lower at 14%.
The actuarial profession stacked up much better when it came to estimating these rates, nailing the median estimate of male smokers exactly at 18%. Our median estimate of a tick under 17% for females was, however, a bit on the high side.
One thing is for sure, I’m not sure I’d want to buy an insurance policy from the company whose actuaries believe that on average 65%+ of males consume two standard drinks and 40%+ smoke on a daily basis.
Or perhaps I would. My premiums might be quite low by comparison.
I pulled together a similar graph to show the range of responses, again split by Australian males on the top (in blue) and Australian females on the bottom (in yellow). On the horizontal axis, I have grouped the responses into 3% bands and highlighted in green where the actual figure lies and highlighted in red where the median estimate lies.
As we collectively nailed the male estimate, I’ve had to get creative and give that block a red and green Christmas theme.
Interestingly, as in Question 1, our female actuarial respondents had a similar overestimate of Australian male smokers (19% compared with 17% for male respondents) but this time also had a higher estimate for Australian female smokers (17.2% compared with 16.5% for male respondents).
Q3: WHAT PERCENTAGE OF AUSTRALIANS DO THINK YOU ARE CLASSIFIED (BY BMI) AS OVERWEIGHT OR OBESE?
Now before we launch headlong into the answers to this question, it is probably worth noting that there are a number of limitations around the Body Mass Index measure (BMI).
In effect, it takes your weight (in kilos) and divides it by the square of your height (in metres) to produce a number. If this number is greater than 25, then you are classified as overweight, greater than 30 and you are classified as obese.
As BMI depends upon the square of height, rather than the cube, it doesn’t take into account the fact that mass ordinarily increases with regards to three dimensions, not two. This means that a larger person will always have a higher BMI than a smaller person, even if their body shape and relative composition was the same.
The medical establishment has generally acknowledged this major shortcoming of the measurement.
BMI also assumes there is no difference between lean mass and adipose tissue – muscle and fat. As a result many elite athletes are often classed as overweight and some even as obese. Furthermore the body fat percentages of individuals in the healthy and overweight ranges can vary widely, with many overweight people having less body fat than ‘healthy’ people.
However, despite all of this, BMI remains a ubiquitous measure of obesity and has been used for 150+ years to provide a ready estimate of an individual’s health.
So, as for the statistics, a higher proportion of males (70%) than females (56%) were classified as being overweight or obese, with the average of 63% increasing from 56% in 1995.
Unsurprisingly, obesity is highest in areas of greatest socioeconomic disadvantage, with reports suggesting this is becoming an increasingly stronger correlation.
Scarily, over 25% of children (between 5 and 18) are overweight or obese.
As for all of you out there, we are wearing seriously rose-coloured glasses. Either that or we collectively failed to account for the broader Australian population. Our median estimate for Australian males was ‘just’ 43% and for Australian females it was 39%. We were well under the actual figures.
Are our life insurance premiums similarly under-priced?
Q4: WHAT DO YOU THINK IS THE LIFE EXPECTANCY OF A NEWLY BORN AUSTRALIAN?
I was quite amazed at the spread of answers that were submitted for this question; particularly those at the lower end of the spectrum. Although, given the levels of obesity mentioned in the previous question, perhaps those respondents are on to something.
At the other end of the scale, quite a few people added in comments like “trick question – mortality improvements” and “allowing for increased life expectancy” to justify their answers. I suspect others just allowed for the improvements; given it was a ‘number only’ input.
Highlighting the ‘wisdom of the crowd’, our range of answers put us within a year of the current life expectancy for both males and females. Males are expected to live for 80 years and females 84 years, according to the ABS, and we expected them to live an extra year in both cases.
SECTION 2 – WAGES AND THE WEB
Q5: WHAT PERCENTAGE OF AUSTRALIANS DO YOU THINK EARN THE MINIMUM WAGE?
The ACTU interviewed a ‘representative sample’ of Australians about the minimum wage. According to the union, the minimum wage is $15 an hour. Which equates to $570 a week or $29,640 a year. 1,360,000 people earn the minimum wage, more than double Tasmania’s population (insert joke about counting people verses heads).
Given that there are 11.53 million people currently employed (as at February 2014), this means that 12% of the population earn the minimum wage.
Although the graphic below shows we have overestimated the number of people that are earning the minimum wage, I strongly believe that part of this is due to respondents assuming the question meant “at least the minimum wage”.
There were quite a number of comments to this point. Worryingly, none of these respondents then suggested anything more than 90%, which would suggest they believe a large number of Australian workers are being underpaid.
Q6: WHAT PERCENTAGE OF AUSTRALIANS DO YOU THINK:
A) ARE DAILY INTERNET USERS
B) USE FACEBOOK
C) ARE AFFECTED BY WEB ACCESSIBILITY DUE TO VISION OR HEARING IMPEDIMENTS?
Well. The internet. Indispensable? Ask the 94% of males between the ages of 15 and 25 that use the internet daily, of which 67% use it after 11pm at night. Given a strong correlation to the 74% of this same demographic accessing the internet in their bedroom or private room, I can only assume they are highly motivated to access the news bulletins for the following day.
Diligent high school and university students, undoubtedly.
According to the ABS, 80% of Australian households use the internet on a daily basis. These figures are a couple of years old, so one can only imagine this percentage would have increased quite substantially. Our median estimate was 76%.
50% of Australians use Facebook, despite some of the accessibility limitations for the vision, hearing and mobility impaired. We were similarly four points under, estimating 46%. Mr Zuckerberg has truly helped to create a behemoth!
But back to the topic of accessibility, 98% of websites failed to meet international guidelines on Web Content Accessibility.
Given that nearly half a million people in Australia are vision impaired and, when including the elderly, over two million are completely or partially deaf. Given the estimated population of 23.5 million, 10.6% of Australians are affected by web accessibility due to vision or hearing impediments. We weren’t very far off this amount, with the median estimate and 30% of respondents suggesting 10%.
The demand for accessibility solutions will only increase as the population continues to age. In fact, in 10 years, the number of Australians aged 65 or over is expected to outnumber those under 14 for the first time in history.
SECTION 3 – HOUSEHOLD EXPENDITURE
In 2013, Australian households spent approximately $650 billion.
Q7: HOW MANY BILLIONS ($) DO YOU THINK AUSTRALIAN HOUSEHOLDS SPENT ON THE FOLLOWING ITEMS?
Item |
Actual expenditure ($) |
Median Estimate |
Cars |
78.4 |
50 |
Public transport |
2.2 |
20 |
Education |
2.0 |
50 |
Recreation |
19.0 |
50 |
Alcohol |
14.1 |
30 |
Tea and Coffee |
1.1 |
15 |
Take away food |
37 |
40 |
National and foreign aid |
5.2 |
10 |
Apparently this was quite a tough series of numbers to estimate. While we collectively struggled to get close to many of the estimates (except take away food!), it was interesting that we felt we spent so much on public transport, education and tea and coffee. However, I suspect that respondents may have taken into account the various tax dollars in the governments’ budgets for education and transport, rather than individual household expenditure.
Out of all those answers, we collectively believed Australians spent more on tea and coffee than national and foreign aid. Fortunately that isn’t the case.
To put our aid expenditure into perspective, Australia gives 35c in every $100 of national income to overseas aid. Total government aid, which is just 1.4% of the Australian Government budget is around five times higher than aid through NGOs.
At $4.50 a cup, our aid is roughly equal to one cup of coffee every week for every Australian inhabiting this country.
Globally, we spend more money on soft drink than total foreign aid to poor countries.
Surely we could do something to increase our foreign aid expenditure.
Just to put some of our estimates into perspective:
- If the cost of the average ticket was $3, every Australian would have to take 280 trips on public transport every year to reach $20 billion.
- At $35,000 for an average new car, the $50 billion we estimated would convert into a new car for each family around every four years.
- $30 billion of alcohol would get Australians around two billion litres of beer at the bar every year. And that’s before we factor in RSL prices!
- At $4.50, every working Australian would have to consume around 1.5 take away coffees every weekday to reach our $15 billion estimate. Imagine the increased productivity – between getting the coffees.
Q8: DID YOU FEEL LIKE YOU HAD ANY ‘REAL’ BASIS FOR YOUR ANSWERS IN QUESTION 7?
14% had some basis for their answers (Some basis), 42% used a bit of ‘actuarial judgement’ (Not really) and slightly more, at 44%, just had a stab in the dark (None). So who got closest? Check out their median estimates in the table below.
Item |
Actual expenditure ($) |
Some basis |
Not really |
None |
Cars |
78.4 |
35 |
50 |
50 |
Public transport |
2.2 |
15 |
20 |
30 |
Education |
2.0 |
50 |
50 |
50 |
Recreation |
19.0 |
50 |
100 |
50 |
Alcohol |
14.1 |
30 |
40 |
30 |
Tea and Coffee |
1.1 |
10 |
20 |
15 |
Take away food |
37 |
40 |
40 |
40 |
National and foreign aid |
5.2 |
5 |
10 |
10 |
Refreshingly, although we may have been quite some way off in some of our estimates, those respondents who believed they had some basis for their answers generally got closer than the other groups. Expenditure on cars was the only one that was further way from the true figure.
SECTION 4 – GOVERNMENT EXPENDITURE
Q9: AUSTRALIANS PAID ALMOST $400 BILLION IN TAXES LAST YEAR. WHAT PERCENTAGE DO YOU THINK WAS SPENT ON THE FOLLOWING ITEMS?
Item |
Actual expenditure ($) |
Median Estimate |
Social security/welfare |
138 |
70 |
Health |
85 |
60 |
Education |
30 |
40 |
Defence |
22 |
25 |
Interest on public debt |
12.5 |
10 |
Immigration |
4.4 |
5 |
All of the information above was taken from the ABC 2013 budget interactive infographic. It is a fantastically insightful tool and I recommend you keep a look out for the 2014 version when they publish it. For those of you who can’t wait that long, I’ve created a link for the 2013 version: ab.co/1gDbv5j.
Overall, I think we faired a little better here than in the previous question. Macro-economics appears to be a little more in our wheelhouse. We almost nailed immigration, defence and interest costs. However we severely underestimated the costs of social security and welfare and health.
Of social security and welfare costs, $55 billion was related to the aged pension. I suspect this may not have been included in many respondents’ answers. Payments to families total $35m, disability payments total $25m and veterans payments are $7m.
Interestingly, and perhaps surprisingly for many respondents, unemployment benefits contribute less than $10 billion to this total.
Out of the health budget, $26 billion was spent on medical services and benefits, $14 billion on NHRPs (state government payments), with an additional $11 billion on the Pharmaceutical Benefits Scheme. Hospital services cost less than $3 billion, while the general administration costs over $3 billion.
Without putting too fine a point on it, the cost of aged care and the aged pension take up a significant part of our federal government’s budget. And there would be few people that would suggest this is going to decrease in the foreseeable future.
This should be quite a wakeup call for many people in our industry looking to develop and regulate retirement income streams.
Q10: DID YOU FEEL LIKE YOU HAD ANY ‘REAL’ BASIS FOR YOUR ANSWERS IN QUESTION 9?
Fewer respondents than in the last questions, just 13%, had some basis for their answers (Some basis). 37% used a bit of ‘actuarial judgement’ (Not really) and 51% just had a stab in the dark (None). So who got closest with Government expenditure? Check out their median estimates in the table below.
Item |
Actual expenditure ($) |
Some basis |
Not really |
None |
Social security/welfare |
138 |
60 |
68 |
75 |
Health |
85 |
40 |
60 |
65 |
Education |
30 |
30 |
30 |
50 |
Defence |
22 |
18 |
20 |
40 |
Interest on public debt |
12.5 |
10 |
10 |
10 |
Immigration |
4.4 |
3 |
5 |
10 |
Unlike household expenditure, those who believed they had little to know basis for their estimates/guesses fared better in the questions about government expenditure. Hopefully none of these respondents are involved in setting budgets in their workplace!
SECTION 5 – AUSTRALIAN EXPORTS
QUESTION 11: WHAT PERCENTAGE OF AUSTRALIA’S TOTAL EXPORTS (BY DOLLARS) DO YOU THINK WOULD BE MADE UP BY THE FOLLOWING ITEMS?
Item |
Actual expenditure (%) |
Median Estimate |
Iron Ore |
18.2 |
30 |
Manufacturing |
12.6 |
10 |
Alcohol |
0.7 |
5 |
Meat |
2.4 |
10 |
Live Animals |
0.3 |
5 |
The next question shows that most people took a bit of a stab at these percentages. However, it is quite interesting how Iron Ore is perceived to be around twice as important to our export dollars as is it actually. There are a number of different minerals, as well as petroleum products, that make up a much larger percentage of our export dollars.
Similarly, alcohol and meat products are nowhere near are important as we may be lead to believe. In contrast, despite talk of competitive advantages elsewhere, manufacturing still makes up over 12% of our exports.
How important are live animal exports to the economy? The ABS data suggests that just 0.3% of all our exports are live animals. 0.3%. Just 0.3%.
Now I like meat more than your average carnivore and don’t have any current desire to separate myself from that particular protein source. However, I struggle to see the benefit of shipping live animals overseas only to be slaughtered shortly after arrival upon foreign soil and no longer under our protection. The month long journeys in cramped, hot conditions will only be examined if mortality rates in sheep and cattle are over 2% and 1% respectively.
We’ve all caught glimpses of the cruelty and inhumane handling of animals, like those broadcast by ABC’s 7:30 Report from the Egyptian facilities during 2013. How can we ever feel comfortable about these practices when we have no authority over the treatment of our animals?
Plenty of ‘facts’ about jobs lost and/or created bandied around between the pro- and anti-live export groups are, in my opinion, a little beside the point when the animals are being treated inhumanely.
It’s about time the Australian population makes a conscious decision as to whether to continue live exports in their current form.
QUESTION 12: DID YOU FEEL LIKE YOU HAD ANY ‘REAL’ BASIS FOR YOUR ANSWERS IN QUESTION 11?
Even fewer respondents than in the last questions, just 7%, had some basis for their answers (Some basis). 34% used a bit of ‘actuarial judgement’ (Not really) and 59% just had a stab in the dark (None).
So who got closest with Government expenditure? Check out their median estimates in the table below.
Item | Actual expenditure (%) | Some basis | Not really | None |
Iron Ore | 18.2 | 30 | 35 | 30 |
Manufacturing | 12.6 | 10 | 10 | 10 |
Alcohol | 0.7 | 2 | 5 | 5 |
Meat | 2.4 | 5 | 10 | 10 |
Live animals | 0.3 | 2 | 5 | 5 |
Finally, the ‘some basis’ respondents have sweeping success, ‘out estimating’ the other repondents, although still being some way from the mark.
CONCLUSION
As one respondent added at the conclusion of their answers, “this was a good quiz… made me think… I know so little!” The purpose of including this wasn’t to make me feel better about myself, despite wading through a barrage of disparaging comments from other respondents, but to serve as a reality check for readers.
What is the basis for your assumptions? Is that basis reasonable? Will it stand up to scrutiny when it counts?
If we are to remain known to be reasonable at estimating reality, we should continue to review and refine our bases.
As one respondent added, wittily, “How often at work do you use the term ‘actuarial judgement’ when it is actually just a stab in the dark?”
a) All the time
b) 50% of the time
c) Rarely or never
d) I might need to use some ‘actuarial judgement’ on that…
CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.