Monday, November 28, 2016

Do the rich always get richer?

Let’s tackle this controversial topic as an example of how thematic maps and simple in-map analysis can bring clarity when dealing with complex issues or situations. 

'The rich get richer and the poor get poorer' is a common catchphrase that is often evoked in discussions about economic inequality. However, is this a true reflection of reality or just a widely held belief that is taken at face value? How do we tell?

It would probably take a few months to fully investigate all the factors that play a role in the complex issue of wealth concentration and to formulate a solid argument which could withstand full academic scrutiny.

However, as often happens in real life, we rarely have the necessary resources at our disposal, or the time, to do extensive research to get to the bottom of the issues we encounter. Therefore, a credible, quickly-determinable clue as to what that answer may be, and which allows us to make a more informed decision, is very often the most practical approach.

Map Deck's newly released Thematic Mapper app - with a unique in-map analytics functionality - offers everyone, even the most novice users, a simple solution that can aid in the search for answers to complex questions. Let’s explore how…

To begin our analysis we need information on the wealth of individuals, as well as statistics on how that wealth is changing over time.

Data on taxable incomes for postcodes for 2003–04 and 2013–14 income years, published by the Australian Taxation Office (ATO) and available for free perusal with MapDeck apps, is a good proxy. This information will give us an opportunity to test whether this widely held belief about rich getting always richer holds true in a spatial context. And in the process, we will also demonstrate how simple mapping tools can provide a reasonable answer - and quickly.

Distribution of taxable incomes (deciles) by postal areas - Adelaide [click to enlarge]

The objective here is to test if top taxable incomes grow consistently at or above the growth rate of taxable incomes of the entire population of Australian taxpayers.

If yes, then this would indicate that wealth of the richest individuals is increasing faster than wealth of the rest of the population and consequently, that poorer people are becoming relatively poorer over time. This is a simple argument but proving or disproving it will be enough to give us a strong clue as to the likely conclusions that would be obtained from more thorough research.

Change in taxable incomes over a decade to 2013-14 - Adelaide [click to enlarge]

It is always a good idea to clearly articulate assumptions and to identify limitations of the data you work with as it is not always possible to have the best and the most accurate information on hand for analysis.

In this case, for example, we defined “the rich” as taxpayers residing in postcodes with median taxable incomes in the top 10% in 2003-4 and in 2013-14.

The limitation of this approach is that we are looking only at the incomes and not the overall wealth of individuals. However, taxable incomes are a reasonable proxy for wealth. That is, wealth usually generates an income stream and even if only a part of that income is attributable to an individual, for example, due to the use of various tax minimisation schemes, growing wealth will be reflected in the growth of taxable incomes of those individuals.

We also assume here that residents of any given postcode are all alike which is an oversimplification but not an unreasonable one. In particular, it is true that postcodes comprise of populations with varying demographic and socio-economic status but summarising their collective characteristics into a single measure is an acceptable practice in data analysis. Therefore, a change in median taxable income for a postcode is a good representation of the change in wealth of its residents.

More of an issue is the fact that we are assuming that the 2016 postcode boundaries are exactly the same as those in 2013-14 and in 2003-04. However, yet again, this is a reasonable assumption since postcodes with the top taxable incomes are predominantly in capital cities and capital city postcodes remain relatively stable over time.

More assumptions and caveats could be listed but the main point here is the importance of being aware of the limitation of the data you are working with. It applies to any data and information, not only to that with spatial context.

It takes just a few moments to set up a map ready for analysis using MapDeck’s Thematic Mapper app. Firstly, we select and import into Thematic Mapper the postcode boundaries and ATO data, then join them using the data mixer functionality.

Thematic Mapper: data mixer

We are interested in only 3 columns from the newly created table: the median taxable incomes for both 2003-4 and 2013-14, and the percent growth of median taxable incomes over the 10 year period. The information on the number of individual taxpayers per postcode is included just for context.

Selecting the “complex legend” option enables advanced data filtering functionality in Thematic Mapper. We set taxable income values to show the top 10% range only (these values are calculated and can be looked up when setting up legend styles for individual data sets from our table).

The last step is to remove empty postcodes from the map (that is, those with no data) and to import a map overlay showing main roads and localities to give our map additional spatial context (the Stamen’s Hybrid layer in this case).

Let’s now review the map, summarise the data and draw some conclusions.

From the map legend we know that:

  • Top deciles of median taxable income were $33,503 to $68,017 in 2003-4 and $51,128 to $113,687 in 2013-14;
  • Median taxable incomes for individual postcodes grew between 7% and 257% over a decade to 2013-14;
  • Median growth in median taxable incomes was 50% (that is, in half of postcodes, median taxable incomes grew by less than 50% in the 10 year period and in the other half the growth was equal to, or greater than, 50% -  therefore the 50% growth rate can be used as a national benchmark for further comparisons). 
Growth in median taxable incomes 2003-04 to 2013-14, top income decile - Perth [click to enlarge]
Growth in median taxable incomes 2003-04 to 2013-14, top income decile - Brisbane [click to enlarge]
The pattern we were looking for (that is, where all top postcodes are coloured in shades of green, indicating growth rates over 50%) is present on both the Perth and Brisbane maps but not on the Melbourne or Sydney maps. Therefore, we can conclude with confidence that it is not entirely true that the rich are getting richer and the poor are getting poorer all the time and in every place.

Growth in median taxable incomes 2003-04 to 2013-14, top income decile - Sydney [click to enlarge]
Growth in median taxable incomes 2003-04 to 2013-14, top income decile - Melbourne [click to enlarge]

In other words, our data and simple in-map analysis indicate that some rich neighbourhoods grew their incomes at less than the national median rate, hence getting relatively ''poorer'' over time with respect to the rest of the population.

Of course, this is only a clue as to what conclusions may possibly be drawn from more thorough research. However, the point is that is takes very little effort to conduct this kind of analysis using MapDeck’s online app and huge range of available data. This approach is not suited for all situations but it can provide valuable insight, and fast, to adequately support many important every day decisions.

Median Taxable Incomes, Australia 2003-4 and 2013-14 map and ATO data table used in this case study are available for free perusal for registered MapDeck users.

And for a limited time, all MapDeck users have the opportunity to try Thematic Mapper at no cost. The app is available for a free one month subscription until 20 December. 

So, what is the big issue you would like to explore next? Use this invite code to sign up if you are not yet MapDeck user: f10dc1f2

About MapDeck:
MapDeck is an online marketplace for spatial information and simple-to-use, task-oriented tools to support a variety of activities, be it business or investment related, environment, community or policy focused.

We make spatial information and analysis ready data accessible to all. Locate, map, act!

Monday, November 21, 2016

Thematic Mapper for easy in-map analytics

The updated Thematic Mapper app, recently released on's platform, unlocks in-map analytic capabilities for organisations of any size or budget.

Traditionally, access to spatial information, and the tools required for analysing and visualising that information, was restricted to only companies with substantial budgets. The two choices for an organisation were to either deploy a suitable infrastructure in-house and employ expert analysts, or alternatively, source specialist advice from external consultants.

Both these options have one thing in common - cost. We are talking about tens of thousands, and very often hundreds of thousands of dollars in annual expenditure. It's no surprise that many organisations could not afford location intelligence at these prices.

The arrival of cloud-based online mapping tools is changing all that, democratising access to vital information for everybody, even the smallest companies, as well as individuals.

Early adopters of cloud-based mapping solutions are realising enormous benefits - both monetary and strategic.

In particular, instant and on-demand access to inexpensive analytical tools and information (on any device) allows individuals at all levels in an organisation to take advantage of location intelligence while making important business decisions. This leads to better outcomes. 

And since better outcomes accumulate over time, it allows an organisation to reach its goals faster, more effectively, and with minimal expense. The benefits of using cloud-based mapping and spatial analysis tools are too good to remain the best kept secret for much longer.

With the release of the upgraded Thematic Mapper app MapDeck is taking the concept of in-map DIY spatial analytics to a whole new level.

Customised thematic map example

Thematic Mapper is an example of the new generation of online mapping tools which allow full personalisation of information content on maps. That is, unlike traditional online maps which require an administrator to configure for end users what data to display and how, Thematic Mapper gives each user full control over the entire map creation and map publishing/sharing process. It is a whole new approach to creating online maps.

Thematic Mapper users can now:
  • filter national and regional data to show just a selection of areas of interest, such as a single sales or franchise territory, retail catchment area, or a local neighbourhood;
  • combine public and private data, such as demographic statistics from Census surveys with company sales and customer data;
  • carry out in-map analysis on data and visualise results on maps and in summary tables;
  • customise presentation of information, from arrangement of preferred base maps and top overlays, to polygon styling according to personal preferences; and
  • publish or share privately all created content and information.

For a limited time, all MapDeck users have the opportunity to try Thematic Mapper at no cost. The app is available on a one month free subscription until 20 December 2016.

Use this invite code to sign up: f10dc1f2

About MapDeck:
MapDeck is an online marketplace for spatial information and simple-to-use, task-oriented tools to support a variety of activities, be it business or investment related, environmental, community or policy focused.

By collecting the most useful spatial data in one place, and providing simple tools to interact with it, MapDeck makes it easy for anybody to take advantage of location intelligence technology and to derive invaluable, personalised information for decision making.

Friday, October 7, 2016

Filtering mapped data by location

When visualising location related statistics on thematic maps – be it the median age of residents, their incomes or expenditure, house price values, or any other statistical information - one of the biggest challenges is how to limit what is displayed on that map to only a specific region of interest, such as a sales territory or similar user-defined geographic area. The updated Thematic Mapper functionality will make such a task a breeze.

This blog post is just another teaser for the upcoming release of the Thematic Mapper app and its purpose is to show off some of the advanced and truly unique in-map analytics functionality that will soon be delivered into the hands of MapDeck users.
Take this very common scenario which demonstrates Thematic Mapper’s distinctive capability: a user/researcher has statistical data covering a whole country but what is needed is only a summary report that profiles a specific geographic region. This area could be an individual sales or franchise territory, an economic region, local government area, or even a single postcode or suburb.

Filtering data on online maps in real-time by location of interest is not an easy task because of the limitations of current online map publishing technologies which require data and map styles to be pre-configured in advance by a publisher. This process cannot be left to individual users.

So, the traditional approach is to cut data into specified geographic areas as the first step in the publishing process and then to use only each subset of data for analysis and for display on the maps.

This is a very time consuming and inefficient process, especially if you have to analyse and publish a few dozens or more area profiles. Not to mention that the process has to be repeated for each user that requires a different set of custom areas (such as sales or franchise territories that each business defines differently). This is the reason why to date  the access to customised spatial information has been so difficult and expensive. 

What the user really needs is the ability to easily “hide” the boundaries outside of the area of interest (that is, mask those with irrelevant data) and to limit the statistical calculations to only a selected subset of data.

Thematic Mapper does this all effortlessly and on-the-fly. Like in the following example, where postal areas have been clipped to a custom geographic region, defined as “Greater Sydney”, and then joined with postal area data from the ABS’ Basic Community Profile 2011 and from the Household Expenditure Survey 2009.

In-map analytics with MapDeck's Thematic Mapper

The resulting map highlights where the largest sub-markets within the Greater Sydney region are and what the total estimated size of that entire regional market is, in dollar terms per annum.

This is yet another example of the power of Thematic Mapper, a truly personalised in-map analytics solution. Sign up to for a free test run of Thematic Mapper when it is released.

Tuesday, October 4, 2016

Bringing the power of in-map analytics to all

The upcoming release of MapDeck’s Thematic Mapper app will make it very easy for anybody to visualise and analyse information on maps. 

We've included below just a teaser to highlight Thematic Mapper’s unique capabilities and to demonstrate the power of a new, totally personalised approach to online mapping.

Like, for example, on-the-fly polygon colouring – in this case, 70,440 residential mesh blocks in NSW by Statistical Area Level 2 region, to highlight actual distribution of housing within each boundary:

ABS Mesh Blocks 2016 coloured by SA2 areas

Or area profiling – in this case, colouring mesh blocks according to land use category and calculating total area of each category:

ABS Mesh Blocks 2016 coloured by land use category

Imagine the possibilities...

Friday, September 23, 2016

ABS releases Postal Areas 2016

Australian Bureau of Statistics has just released a new version of Postal Area boundaries as part of the 2016 update of the Australian Statistical Geography Standard (ASGS).

As the ABS explains, a postcode is a four digit number used by Australia Post to assist with mail delivery. Australia Post does not currently define geographic boundaries for postcodes. However, a number of organisations, including MapDeck, have created geographic boundaries that aim to define the geographic extent of the mail delivery area for each postcode.

Defining postcodes with a geographic boundary is an imprecise process, and this is demonstrated by the fact that there are variations in the boundaries released by different organisations. Additionally, postcodes cover most, but not all, of Australia; for example, western Tasmania is not covered by a postcode.

ABS Postal Areas 2016 coverage

Postal Areas (abbreviated as POAs) are an ABS approximation of Australia Post postcodes and are created to enable the release of ABS data on areas that, as closely as possible, approximate postcodes. This enables the comparison of ABS data with other data collected using postcodes as the geographic reference.

Postal Areas are approximated using one or more Mesh Blocks (MBs) from the 2016 edition of  the Australian Statistical Geography Standard (ASGS). This is a different approach than in 2011 where Postal Areas were built from much coarser Statistical Area Level 1 (SA1) boundaries and had a less accurate representation of actual postcodes.

In developing Postal Areas, each Mesh Block is allocated to a single Australia Post postcode. Postal Areas derived in this way only approximate postcode boundaries. Mesh Block allocations are based on the distribution of the estimated population within each Mesh Block, not on the total polygon area. These allocations have been determined using the best available information on postcode boundaries.

For the 2016 ASGS 2,670 Postal Areas have been defined. The codes used for the 2016 Postal Areas may not match those used in 2011 in some instances. Changes to codes occur where Australia Post abolished postcodes or changed codes between editions of the ASGS.

Map comparing POA 2011 (red) with POA 2016 (blue) boundaries

Some Australia Post postcodes are not included in the Postal Area classification. This occurs when no Mesh Block can be allocated to a particular postcode. There are two situations when this occurs where:
  • a Mesh Block covers more than one whole postcode, the Mesh Block can be allocated to only one postcode
  • more than one Mesh Block partially covers a postcode, but all the Mesh Blocks are allocated to other postcodes, based on the share of population with which they also share area.

Postal Areas exclude Australia Post postcodes that are not street delivery areas. These include post office boxes, mail-back competitions, large volume receivers and specialist delivery postcodes. These postcodes are only valid for postal addresses and are not a valid location for population data.

It should be noted that there are instances where postcodes cross state or territory boundaries.

Unlike MapDeck’s Postcodes 2016 dataset that references only officially gazetted localities, ABS’ Postal Areas are defined to cover the whole of geographic Australia without gaps.

Similarly to all versions of Australian postcode boundaries available on MapDeck, Postal Areas, Australia 2016 are free for use by all subscribers.

Related Posts:
Australian Postcodes Map 2016
Australian postcode boundaries 2016
Explaining holes in Postcodes 2016 coverage

Monday, September 19, 2016

In-map Analytics

Maps are not just pretty pictures - they are data and information visualisation tools that bring out hidden or not so obvious facts about locations of interest.

This one shows where 450,000 jobs are physically located in the City of Melbourne.

City of Melbourne - Employment Distribution 2015

Such information is just trivia for most people but it is an invaluable insight for a business trying to reach this market. Think in terms of finding the best location for a billboard advertisement, strategic parking of company trucks for maximum brand exposure to the largest possible audience, or identifying the best location for opening up a new cafe, etc. 

The ability to mix and filter non-spatial data in real time has been a basic feature of business intelligence software and data visualisation dashboards for many years now. However, the same functionality is very difficult to implement in mapping applications due to the complexities involved in real time merging of spatial data and the information that can be attributed to it.

Traditional online mapping software simply does not allow individual users to select their own data, and create and style maps dynamically to their liking in real time. So there is always a data publisher or administrator involved who decides what can be mapped and how. Needless to say, this uses up valuable time and adds additional costs to the process.

But the age of personalised online mapping, where users select the data then filter it and style however they want, is just around the corner! The upcoming upgrade to MapDeck’s Thematic Mapper app will put into your hands a very powerful, personalised, in-map analytical tool.

City of Melbourne - buildings constructed before 1950.

In the example above, the building footprints (spatial features) and the attribute information (such as the year of construction, building height, number of floors, total area, number of occupants, etc.) come from a different source. Yet, both can be merged easily using standard Thematic Mapper functionality and filtered to show only a required subset of information. And the resulting map can be styled to user liking. No data publishers or administrators are involved in this process.

More announcements are coming soon, stay tuned!

Tuesday, July 26, 2016

In Focus: 2016 Australian Statistical Boundaries

The Australian Statistical Geography Standard (ASGS) defines the boundaries, and respective relationships, of regions in Australia which the Australian Bureau of Statistics (ABS) and many other organisations use to collect, analyse and publish geographically classified statistics. 

To name a few examples, such statistics include Census-derived demographic and dwelling data, future population projections, labour force data and building approval rates.  

ASGS Volume 1 - Main Structure and Greater Capital City Statistical Areas, July 2016 has just been released by the ABS. It deals with the ASGS core structures, such as Mesh Blocks and Statistical Area Levels 1 to 4, and the Greater Capital City Statistical Areas (GCCSA).

The 2016 ASGS boundaries will be used for the publication of the 2016 Census of Population and Housing data and progressively introduced into other ABS data collections.

This is the second edition of the ASGS and it updates the 2011 edition for growth and change in Australia's population, economy and infrastructure. It also incorporates the Territory of Norfolk Island for the first time. Volumes 2 to 5 of statistical geography will be released progressively over 2016 through to 2018.

All boundaries from Volume 1 of 2016 ASGS have been published on MapDeck and complement the previously released 2011 data.

2016 Mesh Blocks on Statistical Area Level 2 boundaries

Correspondence files, describing the relationship between the 2011 and 2016 versions of the boundary data, will be added shortly to coincide with the upcoming upgrade of the Thematic Mapper app. These reference files can be used to recalculate 2011 Census statistics to present them on 2016 version of boundaries. Stay tuned for the updates.