Tax is a huge part of the inequality story
There has been endless discussion about the towering rise in inequality within countries around the world, notably in "Anglo-Saxon" countries. The reasons for it cited are myriad: labour market processes, education, race, gender, culture, technological change, personal preference for work, leisure and risk, and more.
This new study looks interesting. It compares five Anglo-Saxon countries that have relatively similar backgrounds and tax systems: Australia, Canada, New Zealand, the UK, and the US; and it looks at the relationship between taxes and top income shares. We haven't reviewed the paper in detail, but this conclusion is striking:
"The share of the very rich appears to be extremely responsive to changes in marginal tax rates. Over the period 1970-2000, we estimate that reductions in tax rates can explain between one third and one half of the rise in the income share of the richest percentile group."
It also contains a raft of other useful data and analysis, such as graphs showing the change in these shares over time (the graph in this blog is an example.)
As a reminder, outcomes of inequality are important too, of course: note in particular the book The Spirit Level, which demonstrates how a wide array of social outcomes are associated not with absolute levels of wealth or poverty, but on inequality.
This new study looks interesting. It compares five Anglo-Saxon countries that have relatively similar backgrounds and tax systems: Australia, Canada, New Zealand, the UK, and the US; and it looks at the relationship between taxes and top income shares. We haven't reviewed the paper in detail, but this conclusion is striking:"The share of the very rich appears to be extremely responsive to changes in marginal tax rates. Over the period 1970-2000, we estimate that reductions in tax rates can explain between one third and one half of the rise in the income share of the richest percentile group."
It also contains a raft of other useful data and analysis, such as graphs showing the change in these shares over time (the graph in this blog is an example.)
As a reminder, outcomes of inequality are important too, of course: note in particular the book The Spirit Level, which demonstrates how a wide array of social outcomes are associated not with absolute levels of wealth or poverty, but on inequality.
posted by TJN at 2:38 am
3 Comments:
This would rely, however, on the assumption that all income is declared. The reality may be that it for the top quintile it is income declaration, rather than income, which is highly sensitive to the top marginal rate...
Following on from Alex's comment on undeclared income, a similar issue applies to undeclared wealth. Statistics on personal wealth already show a massive concentration of wealth in the hands of a tiny elite in most countries. But do the statistics tell the whole story? Judging from discussions with World Bank officials and others, wealth statistics do not capture wealth held through legal structures, including trusts, offshore companies and such like. This being the case, levels of inequality are likely to be much, much higher in most countries than the already appalling position described by the official statistics.
Correlation does not imply causation. Repeat 100 times!
"The share of the very rich appears to be extremely responsive to changes in marginal tax rates." There is an implied thought process here that tax rate is the stimulus and wealth share is the response of the system. But there's loads of explanations for an observed relationship like this. Most obvious:
A) Reductions in marginal rates cause the rich to take a higher share of a given economy.
B) The very rich are more successful at causing marginal tax reductions the greater the share they have.
C) The very rich migrate away from places with high marginal rates of tax.
D) High marginal rates harm the economic success of the most wealthy and so shrink the economy
The problem is that these different interpretations suggest different policy responses. (D), for example suggests lowering marginal rates, whereas (A) suggests increasing them (assuming a bigger, more evenly distributed economy is better).
So the data in the paper don't give guidance as to how to improve the tax system. The right approach is via economic modeling (simulation) and testing.
Post a Comment
<< Home