« "Progress" | Main | Our illiberal culture »

September 14, 2011


Feed You can follow this conversation by subscribing to the comment feed for this post.

Duncan Stott

But the Laffer Curve only works for government revenue. The curve for GDP or well-being could be any shape, e.g. it could be something other than 0 at 0% or 100% tax rates.

Paolo Siciliani

And also, I don't like this defeatist attitude towards people eluding taxes. Greece and Italy have been brought to economic breakdown also because of the wide perception among the public that there is no way to rein in tax dodgers. There is always a way, what is needed is just to political will to act. Cynicism is exectly what the tax dodgers want......or technically speaking, there is endogeneity in your first Laffer curve.

Left Outside

Well, there's actually six...

- does it maximize revenue in the short run?
- does it maximize revenue in the long run?
- does it maximize GDP in the short run?
- does it maximize GDP in the long run?
- does it maximize well-being in the short run?
- does it maximize well-being in the long run?

The long run maximum tax revenue raising rate will be identical with the long run GDP rate, and there's an argument it will align with the long run well-being rate.

In the short run these curces will likely be different, but over the long term, they are likely to converge.

There are also dynamic factors to consider. A low rate may reduce the payoff to lobbying for the introduction of loopholes. Since loopholes complicate the tax code and make tax dodging easier they move the laffer curve to the left. Setting the short run tax rate at its laffer maximum might set in process a diminuation of the maximum long run laffer curve.

Dominic Arnett

Unfortunatly the only way to test theory is to do it and this can lead to disaster in the economy. This leads to slight changes in policy rather than big changes.

The Pedant-General


Your third example (reducing GDP, raising well-being) hints at another bit of tax philosophy - you want to tax undesirable behaviour, and encourage (or not tax) good/desirable behaviour.

In these cases, the aim of the tax is NOT to raise revenue, but to deter the undesirable activity.

I struggle to see that the 50p rate on income tax adheres to either of those conditions:
- the aim is explicitly to raise revenue, so if it isn't doing so, then what is its point?
- you don't (normally) want to treat work as an undesirable activity.

Luis Enrique

do you really think, on net, high taxes are a good way to deter "undesirable activity"?

I mean OK, you can think of some potential examples, but you can also think of examples where it deters desirable activities. What matter is the whole distribution, so to speak.


@ Luis, PG. There is a reason to think that high taxes might effectively deter "bad" economic activity.
It's that the motives for good and bad activities differ. A genuine entrepreneur might be driven by intrinsic factors - the desire to create something. But a rent-seeker is likely to be more money motivated, and thus more deterred by a lack of financial reward. To the extent that there is this difference, then high taxes might well be effective.


@ Luis, PG. Perhaps less about undesirable activities and instead simply about the impact on income distribution? High marginal taxes on higher earners would flatten income distribution, which in turn could increase society's well-being.


To focus only upon the first of these, the revenue effect...

Tax - definition
a. A compulsory contribution to the support of government, levied on persons, property, income, commodities, transactions, etc., now at fixed rates, mostly proportional to the amount on which the contribution is levied.

Nothing here about happiness - all about revenue.

Luis Enrique

Yes, but what about people who merely do jobs that the market places a high wage upon, neither entrepreneurs nor rent seekers? Doctors, accountants, whatever. That's what I mean by needing to think about whole distribution.

Paolo Siciliani

Anyway, I doubt that higher tax rate might even deter "the sort of activity that is negative-sum for the economy as a whole".

To the extent that these people care about their relative status, they will work hard up until their life style is conspicuously superior than the average worker regardless (or better still in spite) of the marginal tax rate. Whereas, my understanding of why the Laffer curve would reduce revenue is that people pursue an "absolute" level of well being, hence they are put off from working harder if you raise taxes.....

If this makes sense, provided that you cannot deter these sort of undesirable activities through the tax system (since it's root cause is sociocultural), it is certainly better to raise taxes as a way to increase revenue. So, under these circumstances, higher taxe rate would lead to an increase in revenues and (alas I suspect) an increase (decrease) in GDP (well being) as these status seekers will work harder to satisfy their expectations....


«it could be that the tax rate that maximizes revenue is quite low - because even a modest tax will lead to tax-dodging»

This statement is of such staggering naivety.

Raising tax is not dependent on one «tax rate» because a crucial aspect of a well designed tax system is that it makes avoidance and evasion difficult.

Therefore a multiplicity of taxes raising a bit of revenue from fairly many sources is desirable, because it reduces the incentive of cheating any one of them.

As such the 50% tax is not a big deal. Compare to the 20% VAT tax that the poorest have to pay on a very large part of their spending.


«another bit of tax philosophy - you want to tax undesirable behaviour, and encourage (or not tax) good/desirable behaviour.»

That's a crazy ta xphilosophy, and while conservatives seem to think that tax should be raised on the poor to discourage idleness and parasitism, and decreased on the rich to encourage virtuous production, that's not what tax is for, as it is not a morality tale.

The primary purpose of tax is to raise revenue (or control inflation), very much because it has to be raised, not because it is a good thing to tax people, whether more because they are exploitative poor or less because they are deserving rich.

So a tax system should be designed to just raise whatever overall revenue has been politically decided, in ways that are:

* Ideally neutral.
* Relatively cheap.
* Expensive to avoid.

That's just about all. Then the details can deviate from that because of politics; for example it is much easier to tax low income people via indirect and direct taxes than to tax rich people via income or capital gains and other direct taxes, so the tax system tends to squeeze harder the poor than the rich, but that is not necessarily related to pushing the craven poor and rewarding the virtuous rich; it is just that the poor have less resources to invest in tax avoidance and evasion.

The particular reasons for the increase of the tax of incomes over 150k by 10% is that it is rather easy to do, and it is a very modest increase that is well below historical levels that proved to be quite bearable.

There is an element of politics in that a lot of people are paying for the recession by being unemployed or having big income cuts, and people earning above 150k obviously are not as bad off as others, so if borrowing has to be contained, perhaps it should not be raised by raising VAT or cutting too much unemployment insurance, however much the unemployed may be regarded by some as idle scroungers living in luxury at other people's expense, and people most affected by VAT increased be despised as unproductive losers.

George Carty

Wouldn't an extremely high tax rate on very high incomes (say 90% for incomes of over £5 million per year) be a good idea even if it doesn't raise any net revenue, because it would destroy the incentive for greedy corporate bosses to pillage the corporations they control for their own personal gain?


George, yes. That was the main aim of the original 90% income tax on the top 1%:


The comments to this entry are closed.

Why S&M?

Blog powered by Typepad