Green dawn rising?

Posted by Joshua Newton On September 29, 2010 0 comments

The Greens had a particularly good showing in the 2010 Federal election, winning their first lower house seat in a general election, receiving their highest ever share of the primary vote, and increasing their Senate representation to nine. Concerns have been raised in the Victorian state branch of the ALP, however, that these gains came at the expense of the ALP. So, how seriously should the ALP be concerned about the Greens?

To examine this issue, I created state-specific scatter plots tracking electorate-level results from the last four Federal elections. In each plot, the vertical axis represents the percentage of primary votes cast in favour of the Greens while the horizontal axis represents the ALP two candidate preferred vote.

Click on the image above or here to view the plots in full size.

As the plots indicate, the Greens primary vote has been rising since the 2001 election. These increases have occurred irrespective of whether the ALP two candidate preferred vote has trended towards the ALP (Victoria, South Australia, Tasmania, ACT, NT), away from the ALP (Western Australia), or moved in every-which direction (New South Wales, Queensland). As such, there does not appear to be a consistent relationship between the Greens primary vote and the ALP two candidate preferred vote.

So, should the ALP be worried? Yes and no. In most electorates, the Greens are unlikely to threaten ALP candidates, at least in the foreseeable future. The exceptions are in inner-city electorates held by ALP MPs. In 2010, for instance, the Greens primary vote exceeded that of the Liberal Party in three electorates (Grayndler, Melbourne, and Batman), forcing a run-off between the Greens and the ALP. One of these electorates (Melbourne) ended up going to the Greens, and future gains in the Greens primary vote could also see the seats of Grayndler and Batman come into contention.

  • An explanation of the two candidate preferred vote can be found at the AEC website.
  • Four electorates (Kennedy, Lyne, New England, and O’Connor) were not included in the plots because the ALP was not represented in the Two Candidate Preferred Vote.

The environmental impact of a roast dinner

Posted by Joshua Newton On September 12, 2010 0 comments

Roast lamb and roast beef are classic Australian dinners, and both are, in their own right, quite delicious. In recent years, however, there have been a growing number of articles in the popular press examining the environmental impact of eating meat. Which got me thinking: is it more environmentally friendly to eat lamb or beef? And how do these meats stack up against the other ingredients that are commonly used in a roast?

The data presented in the graph below relates to the amount of water (in cubic metres) and CO2-equivalent emissions (in tonnes) required to produce one tonne of the selected ingredients in Australia. Lamb appears to be more environmentally friendly than beef, at least when evaluated in terms of water and carbon emissions. Nevertheless, both meats have a substantially greater impact on the environment than vegetables. So, next time you are making a roast, think about using lamb instead of beef and consider reducing the amount of meat that you use.


  • Yes, pumpkin and onions are essential elements of a roast, but I couldn’t find Australian water consumption figures for these vegetables.
  • The Peters et al. (2010) paper provides several estimates of the CO2-e emissions associated with both beef and lamb production. The means of these estimates were used in the graph above.
  • The CO2-e figures relate only to the production of the raw ingredient. Thus, the energy required to transport and prepare these ingredients is not considered.

International test cricket, 1971-2010

Posted by Joshua Newton On August 14, 2010 0 comments

During the early 2000s, I probably watched most of the international test cricket matches played on Australian shores. Perhaps unsurprisingly, this period also coincided with Australia’s dominance in the world of test cricket. Australia had not, however, always been so successful. In the 1980s, for example, the West Indies were the most dominant international team and the Australian side enjoyed only sporadic success. Similarly, in the last couple of years, Australia has had fewer series wins than India or South Africa. Which got me thinking: how could I best represent the cyclicality of test cricket success? The graph below is my attempt.

In this graph, I’ve only included test sides that have been ranked number one at least once since 1971. Thus, New Zealand, Pakistan, Sri Lanka, Zimbabwe, and Bangladesh are not included. The height of each arc represents the number of consecutive test series that each side won (or lost), and the spacing between each arc represents the length of time that this run of wins or losses lasted. Flat lines (i.e., when no arcs are present) refer to drawn test series.

Click on the image above or here to view the graph full size.

  • South Africa was excluded from international cricket during the apartheid era, re-entering the international test arena in 1992.
  • Official ICC rankings were introduced in 2003. The unofficial rankings presented here were calculated by Dave Wilson.


The geography of health

Posted by Joshua Newton On August 1, 2010 0 comments

On the 31st of August, 1854, an outbreak of cholera struck the Soho district of London. At the time of the outbreak, the prevailing theory was that cholera and other infectious diseases were spread by ‘bad air’. Dr. John Snow, a local physician, was sceptical of this theory and, with the help of a local minister, found sufficient evidence that cholera was a water-borne disease to convince local officials to disable the water pump at the heart of the outbreak.

In subsequent years, Snow went on to map the residence of each person who was infected with cholera in the 1854 Soho outbreak. The key insight from his map was that disease – and more generally health – has a spatial dimension. In the case of Snow’s map, this spatial information provided some clue as to the cause of the health condition. In many contexts, however, simply seeing geographical variations in health-related indicators can be of interest, especially to public policy makers.

Each of the maps below represents the local statistical areas that correspond to the suburbs of Melbourne, Australia. As can be seen from the maps, certain health-related indicators vary dramatically across suburbs. Notice in particular the greater rates of physical inactivity and psychological distress in the inner western and outer south-eastern suburbs. Also note that smoking rates appear to increase with distance from the Melbourne CBD.


Per capita CO2 emissions

Posted by Joshua Newton On July 22, 2010 0 comments

Discussions about mitigating the effects of climate change invariably lead to national comparisons of greenhouse gas emissions, and such comparisons reveal that China is the world’s largest emitter of CO2, followed closely by the USA. One of the problems associated with such comparisons, however, is that they fail to account for national differences in population. A fairer comparison would therefore be to examine per capita CO2 emissions.

The graph below contains the per capita CO2 emissions of various nations in 1990 and 2007. As the graph reveals, many nations throughout this period experienced dramatic growth in per capita CO2 emissions, and this growth was particularly evident in the Asia-Pacific region. It should be noted, however, that much of this growth took place off the back of low per capita CO2 emissions.

Juxtaposed with this growth was the small to moderate declines in per capita CO2 emissions throughout Western Europe and the former Soviet Union. Nevertheless, per capita CO2 emissions in many of these nations still remained high. If we are to take to heart Jefferson’s sentiment that ‘all men are created equal’, then we must move towards a future where everyone has the same right to emit the same (sustainable) amount of CO2.


Graphing the Tour de France

Posted by Joshua Newton On July 14, 2010 0 comments

The Tour de France has just entered the second week of racing, and I’m tired. In a good way. Because if you’re an Australian fan of the Tour de France and you like to watch each stage live, you must be prepared to lose a little shut eye.

While I love the TV coverage of the race, it can sometimes be difficult to gauge how cyclists who are not race favourites fare on each particular stage. To this end, I created a graph using data from the 2009 Tour de France to determine the stages that had a major effect on cyclists’ times.

Click on the image above or here to view the graph full size.


Obesity in the USA, Part 2: Fast food

Posted by Joshua Newton On July 9, 2010 0 comments

In an earlier post, I showed some of the macro trends associated with obesity in the USA. In particular, I showed how increases in the price of fruit and vegetables relative to the price of soda and sugar were associated with rises in BMI and energy intake. In this post, I take a more focused look at the issue of obesity in the USA.

The graph below shows how the maximum serving size of several products offered by McDonalds and Burger King has changed over time. What is particularly noteworthy is that the amount of soda currently served to children at McDonalds is larger than the largest serving size available in the 1950s. It is also apparent that since the 1950s, the maximum serving sizes available for soda, French fries, and hamburgers have increased dramatically. Food for thought!