Wednesday, May 16, 2012

Stirring the data with an urban spoon



One of life’s great tragedies is the limited opportunity for meal consumption, usually estimated at 3 per day.  The seriousness of such a condition is accentuated when travelling through cities at fast pace.

Having failed to find an adequate pizzeria on the first of a two-day fly by Napoli, the alleged birthplace of such fine cuisine, the pressure mounted on my second day.  Without the three-meal-per-day phenomenon I could have tried every pizzeria I walked past, but knowing I would only get one crack at it, the choice would define my soon-to-be memories of this wonderful city.  The three-meal standard was once again proving a major limitation on my savouring life. 

After hours of walking, pizza napoletana.
I walked up and down the Neapolitan streets in search of this fabled slice, finally giving up as the sun was setting, mostly due to hunger, anger, and my train’s imminent departure.  In a final bout of desperation I succumbed to the type of ‘cheesy’ touristy pizzeria I would otherwise arrogantly lift my nostrils at.  Not even the world’s best table red (coca cola) could help me wash down the shame and disappointment.



Back in home territory the opportunities to try the local cuisines might not be so limited (numerically speaking), but the desire to maximise every outing still holds.  Over the years I have realised that a restaurant’s reputation is not a great predictor of satisfaction.  Neither is the size of its queue. But perhaps among the masses lie some nuggets of truth for us to take away. 



Websites such as urbanspoon provide peer reviews rather than the “hat” dispersing scriptures dividing good food from evil.  Usually used to check out restaurants individually, the data held by these online voting booths can surely tell us more about our candidates.



While not including every restaurant in Australia, the 22,297 profiled in capital cities across the country appear to constitute a majority of the population.  According to a quick search in the Yellow Pages there are 31,900 “Restaurant and Cafes” across the capital cities, and an ABS publication estimates the figure at 15,423 in 2007.  While the scopes and definitions might differ slightly, the figures suggest it's a decent sample.



When looking at restaurant reviews in bulk a few things become apparent:
    Reviewers are generally positive creatures.
    Cost driven expectations are not matched by experience.
    If you love your steaks, eat at home.


Urbanspoon’s rating system is expressed as a percentage of the people who liked it vs. those who disliked it.  On average restaurants get liked 78% of the time.  This average "like score" is consistent (+/- 3%) across all capital cities*.  While almost 40% score in the 80s, less than 5% of restaurants have scores under 50%.






Surprisingly, these positive reviews are less likely for the expensive restaurants than el cheapo ones.  The elevated expectations which come with paying extra seem to lead to disappointment.  


Overall, cheap restaurants score 9 percentage points more than expensive ones, with the middle ranges falling in between.  This trend occurs not only across all capital cities*, but also across most cuisine types.  Cheap Thai beats expensive Thai, but then again cheap Thai beats pretty much everything!  


French restaurants (and to a lesser extent Japanese) seem to disturb the pattern.  Of the cuisines with large enough sample sizes, les French are one of the only cuisines not to show decreased satisfaction likelihood with increased price, keeping a steadily high level across all price ranges (80% like).  

So why pay more if we’re less likely to like it?  The difference in price is not insignificant, so if these data are anything to go by: cheap is cheerful, and pricey just makes disappointment more likely.  

Sometimes we don’t crave fancy or convoluted: forget the foam and the parfait, the cocktails and pate. Sometimes we just want a bit of steak (aged, of course, wagyu if possible).  In that case, do not head to a steakhouse.  Of all categories of restaurants on urbanspoon, steakhouses fare the worst with an approval rating of just 69%!  


How hard can it be? 


Perhaps it's because it doesn’t come with the scent of closely mowed grass, but as simple as a steak is to serve (usually with a side of chips and mushroom sauce) steakhouses just don't cut it.  

On the other hand vegetarian restaurants score 80%, up there with Turkish, Korean, and Fish and Chips!




______________________________


* When discussing capital cities, Hobart, Darwin and Canberra are excluded due to their small sample size.

Friday, February 24, 2012

Two-speed tax decrease

Seeing as they are one of life’s two certainties, how is it that we're so ignorant about taxes?

The federal government updates the income tax system roughly every second year, more often than not announcing tax cuts.  This kicks off a war of sound bites over who is prepared to deliver the deepest cuts and for whom.  What they rarely provide, however, is a clear and simple indication of what the impact of each tax plan will be on one's paycheck.  While providing the intended tax brackets allows people to conduct the necessary analysis, I doubt most people would bother to do so, thus making assumptions on the impact of the changes.

Since converting the national currency to the Australian dollar in 1966, the average income has increased from just under $3,000 dollars per year to almost $55,000. Although the rate of growth has fluctuated, it has constantly been positive, with no year posting a negative growth.
*There is a break in series in 1981, as figure prior to 1981 are "Average Male Income", and figures since 1981 are "Total Average Income"




Due to the combination of Australia's progressive tax system and the continuous income growth, a government's failure to update the tax system quickly translates into a tax increase.  For example, a person earning $50,000 this year is likely to earn around $51,500 next year due to income growth (average growth of around 4% per year since 1990).  This growth puts a larger percentage of their income in a higher tax bracket than before, therefore increasing their overall tax rate.


The graph below shows what would have happen to a person earning the average wage every year, had the tax rates been frozen in 1990-91.




Although the difference is small from one year to the next, a twenty year time-line clarifies the point.  While the average income in 1990 was $25,910 (falling within a tax bracket of 38.5%), 2011’s average income was $54,700, attracting the top tax rate of 47% back in 1990.





Income tax is a complex beast with many aspects including levies and rebates changing year to year.  In an attempt to simplify the history of income tax in Australia, the following graph displays the percentage of tax paid for different multiples of the average wage from the inception of the Australian dollar onwards.
*The resident Prime Minister is only indicative, as their timing doesn't line up nicely with the financial years.



Other than a very curious bump in the late seventies*, the years since have not provided much excitement. The slow and gradual "ramp-up" through the nineties is explained by almost a decade of non-updated tax rates, similarly to a small bump in the early to mid 'naughties' (2001-2004).  Otherwise it has been a steady decrease of income taxes since Malcolm Fraser's first term.  


Generally speaking, income taxes have decreased over the past few decades.  This decrease, however, is not evenly spread across the pay brackets.  While the average wage now gets taxed about 4 percentage points less than thirty years ago, people earning 3, 4 and 5 times the average wage have experienced decreases of 9, 11 and 12 percentage points.  This trend is even more pronounced if we used 1985-6 as the base year.  Since Hawke's first term in office, income tax on the average wage has decreased 2 percentage points, while those earning over 3 times the average wage have experienced a 14 percentage point drop in their income tax!


Interestingly, people earning half the average wage are still paying roughly the same levels of tax as they did in the late 70s, hovering around the 12%.  


Of course, none of this includes the impact of tax breaks for Low Income Earners, nor does it include negative gearing or other benefits offered through our tax system.  It is purely a simple model of calculating income tax percentages, by different multiples of the average wage.  

The picture it tells appear lopsided, though mostly because there seems to be no other side in modern Australian politics.



_________________________


For references:
Tax brackets were sourced from the Australian Taxation Office's (ATO) website:
http://www.ato.gov.au/individuals/content.aspx?doc=/content/73969.htm
Average Income from the ABS' Average Weekly Earning's publication:
http://www.abs.gov.au/Ausstats/abs@.nsf/0/14CDB5CD59F6A075CA2575BC001D6157?OpenDocument