Tuesday, 13 June 2017

The end (of the Census) is nigh

I am unsure whether that headline means that
  • the end of the 2016 Census cycle is approaching, with first "real"results due to be released on 4 July 2017; or
  • the 2016 exercise will prove to be such a fiasco that there won't be another attempt at a full count.
The announcement that drew my attention to the first of these was the release of shells (I presume empty tables) for Table Builder.  This led me to revisit the 2011 data with mixed success leading to the second possibility.

Following my posting about the imminent release to the Community Facebook page I had a couple of comments that got me trying to compile a bit of analysis comparing Carwoola, Captains Flat and Bywong/Wamboin.  My first foray (an age profile) was quite successful but when I tried again (an even simpler table on sex) it just basically wouldn't work for 2 days (Saturday and Sunday).  It did work on Monday 12 June, although possibly a bit slower than in the past.  Breaking news: I attempted to download an age*sex table for Australia on Tuesday 13 June and while it loaded the shell data quite quickly the data retrieval for this very simple table had not completed after 18 minutes so I killed it.  Interestingly by 6pm on the 13th it was working well again.

My thinking  about this is that there was so much demand, except on the public holiday, on the servers that the ABS system couldn't cope.  The same problem as for the Census Night shambles.  Comment has been sought from ABS.

Here is the information I have gleaned.  (The BIG PROBLEMO  with this sort of undertaking is that as one looks at basic information and tries to understand there is always "one more cross-tabulation that will explain an interesting number" so one digs deeper and deeper and eventually disappears where the sun don't shine.  I have tried to maintain a little discipline in this.)

Overall Population

Bywong/Wamboin 2,803
Carwoola                1,383
Captains Flat             723

The relatively large size of Bywong/Wamboin means that 'they' tend to dominate any aggregate of the three areas, so I have generally kept them separate.


The most interesting differences are between the Flat and the other areas with a higher proportion of 0-9 year old children (and less 10-19 year olds) and  a higher proportion of  30-39 year old  adults.
Bywong/Wamboin is particularly deficient in 20-29 year olds.


Having this as a heading should increase my click-count, but if you came here looking for monkey business I'm afraid you're out of luck.

The traditional way of presenting the balance between the sexes is the ratio of men per hundred women.  The standard demographic belief has this as 105 at birth, after which it drops - quite rapidly after age 20 - as blokes do what blokes do (often involving motor vehicles and/or power tools) reducing the age specific rates below 100.  Here is a sample graph for Australia (from the 2001 Census because Table Builder initially refused to deliver the 2011 stuff).
The overall ratio for 2001 was 97.4.  Eventually Table Builder gave me information for 2011 showing the overall sex ratio to be 97.7 - effectively the same.

For the three areas I am interested in the overall ratios were:
Bywong/Wamboin  95.5
Carwoola               110.8
Captains Flat         111.1
While the value for Bywong/Wamboin is close to the Australian average those for Carwoola and the Flat are quite high.

My next step was to cross-classify the two items.  In view of the small populations in the study areas using single year ages would be subject to considerable instability due to small absolute numbers. Looking at the various series showed  that a major contributor to the high masculinity of Carwoola was the 20-29 age group with 95 males and 47 females.  This led me to create the following three charts showing the sex ratio together with each age/sex cell as a percentage of total population of each area.
 I find it hard to rationalise this information but ;looking at the shapes of the various lines it seems clear that the reason for the high sex ratio in the 20-29 age group is the low number of females.  I offer no suggestions as to why this might be but would welcome ideas from others.
 In contrast to the Carwoola situation both sexes leave Bywong/Wamboin in their 20s.  I have a glimmer of an idea about people in urban areas buying a house in the area where they grew up, and there is enough churn in the housing market to allow that.  Out here there are relatively few houses coming up for sale (and not that many employment prospects) so folk move elsewhere.
Again very few young adults in Captains Flat.  The sex ratio is very high for kids.

Marital status

This is the third of the "basic demographic" topics.  There is little outstanding in this graphic, although more folk in Captains Flat seem to be in de facto relationships than the other areas.

Household Composition

For the Census much weight is given to the concept of "household" which means, in concept, people in a dwelling who eat "from a common pot".   My experience is that this is a rather outdated concept, especially for group households.  However in our area, such niceties don't really matter as the vast majority of dwellings are occupied by a single family.
The 'Other not classifiable' category is mainly those households which the Census collector determined were occupied on Census Night but where the Census collector could not make contact. 

The more interesting (starred) items here are:
  • The very high percentage of single family households in Bywong/Wamboin;
  • The very high proportion of lone-person households at the Flat; and 
  • The relatively low proportion of "Other non-classifiable" households in Bywong Wamboin (the collection staff there were either 
    • very good and made contact with nearly all occupied dwellings; or
    • very bad and classified all non-contacts as vacant!)
At this point my good intention about not chasing random ideas became a tad inactive as I decided to compare the household type variable for the total of the three local areas and Australia.
I have starred the areas that appear to have a large difference between our areas and Australia.  I find it particularly interesting that the data appear to confirm a belief that it needs a family operation to manage a rural-residential property!  Very few lone person (despite the Flat's efforts) even lower number of group households.

Showing that not all discipline has vanished I have not attempted to explore the situation of two family households (yet - I may have a go later).


Anonymous said...

Re: age profile, the availability of a primary school and preschool clearly influences Captains Flat (and conversely the lack of them affects the other areas). When the kids reach high school age, there's an incentive to move closer to "town", mainly Canberra/Queanbeyan, especially if there are evening and weekend sports that the parents have to drive the kids to (school buses taking care of school transport, but at the loss of hours per week to the kids).

Small house and block sizes in Captains Flat probably influence the household size, plus the lower values making them more viable for single-income households.

Peter M

Flabmeister said...

Thanks Peter.

Flabmeister said...

From Sandra (through Facebook) You have suggested that the high proportion of family household groups reflects the fact that it takes a family to run a rural/residential operation. It might instead mean that the country lifestyle is conducive to happy families which stay together!!

Flabmeister said...

From Cathie and Galle via Facebook Primrose Valley is always forgotten!

My response on Facebook: There are several issues here. 1 Oops. Fair call. 2 What I have done in the past is include Primrose Valley, Carwoola and Hoskinstown together as "Gazette catchment" but given the problems I was having with Table Builder decided to keep things simple. 3 The suburb of Primrose Valley as included in Table Builder crosses the Queanbeyan River and includes most/all of Urila!!!! I have raised this with ABS in the past but am not confident they'll have sorted it. I hope the Collector has a snorkhel on their vehicle.