Mar 05, 2012 11:00
music,
business,
visualization,
broadcasting,
astronomy,
google,
language,
scotland,
society,
women,
cat,
feminism,
falklands,
france,
bechdel,
edinburgh,
sex,
reddit,
radio,
cooking,
possum,
sexuality,
democracy,
links,
history,
warcrimes,
photo,
cute,
uk,
media,
funny,
comics,
video,
intelligence,
space,
gender,
agriculture,
photos,
cartoon,
housing,
food,
lgbt,
politics,
gaelic
I have a maths degree, regularly attended seminars in economics (as they have a high maths content) and conclude that while economics is a fascinating subject and well-founded mathematically, the evidential basis for most economics is not so clear cut that we can yet make important decisions on that basis.
Take two really basic "macro economic" questions:
1) At which point by raising the tax rate do we lose money in the short term by "rich people leave".
2) At which point in the taxation curve can we reduce tax rates and get more money in the longer term (because people are incentivised to work harder by the lower taxes) -- see Laffer Curve and Reganomics.
These are really very basic questions about economics but have no clear cut answer from mathematical and statistical answers yet.
So, despite having strong opinions on economics myself, I'm not convinced too much mathematics helps. I do worry that people tend to default to "simple" arguments -- e.g. "low tax is good because you have more money" is a simpler argument than "with higher tax we can increase the productivity of society as a whole by pooling our resources to increase net wealth and money invested as tax often produces returns which outstrip the investment so by paying £1 in tax you often accrue much more than that £1 in reward although indirectly".
I believe the second argument is much more correct but the first argument is much easier to make in a debate where the people judging are of average intelligence.
Reply
Leave a comment