mnot’s blog

Design depends largely on constraints.” — Charles Eames

Friday, 2 July 2004

Geopolitical Arbitrage

Filed under: Economics

To develop a previous theme;

As markets become more transparent and liquid, it becomes more easy to see opportunities for arbitrage. A wide-scale example of this (to stretch the definition of “arbitrage” a bit) can be seen if you consider the growing ease of moving between countries, and therefore taxation regimes, due in part to globalisation.

In other words, free trade affects jobs not only by moving jobs, but also by moving people, and when people have a choice about where they live, one of the things that affects that choice is taxation and the benefits it brings, like education and health care.

In other other words, where can I move to get the most bang out of the least tax buck?

Take One

I’ve seen other comparisons of taxation systems, but they seem to focus more on the associated incentives and disincentives to business (e.g., payroll taxes). To my knowledge, there isn’t much comparison of taxation from the perspective of the taxed. I’m not interested in the macro, economy-wide effects of taxation; I want to know about the personal effects (which, in aggregate, will have economy-wide effects).

To answer the questions I have (such as, “What lifestyles are countries trying to promote with their tax policies?” and “What countries have the most progressive tax policies?”), I’ve started putting together an Excel spreadsheet that tries to compare countries’ taxes and benefits on as much of an apples-to-apples basis as I can*.

This is possible through the magic of of Purchasing Power Parities, or PPP. Without going too much into the details, PPP allows you to adjust currencies for the prices of common goods and services, such as food, furniture and doctors’ visits, thereby giving a truer conversion of a currency (for some purposes, at least) than exchange rates. Perhaps the most well-known and evocative illustration of PPP is the Economist’s Big Mac Index.

PPP, in combination with a painstaking consideration of national and local income tax rates, deductions, credits, costs of social security and healthcare, allows one to paint a fairly realistic picture of the comparative costs in different locales. This means, for example, that in the US, the cost of healthcare is included, because in many other places, it’s part of the package you get for your taxes.

To start with, I’ve selected four locales; California in the United States, Australia (which doesn’t have much in the way of local taxes on income), London in the United Kingdom, and Stockholm in Sweden.

The spreadsheet (Excel) works by allowing you to change the household makeup (number of adults and children), income mix if there are two adults, amount of retirement savings and amount of mortgage interest paid, seeing the effects for each country across a broad range of household incomes, in PPP dollars.

These calculations are very rough; for example, both the UK and US figures don’t take into account many tax deductions and credits that will offset much of the lower end’s tax. Also, retirement tax benefits for the UK and Sweden aren’t capped yet, so the graphs for them may be misleading with certain settings.

In a future revision, I’d like to take account of capital gains taxes, child care and education benefits and expenses (if applicable). Of course, it would also be good to add more countries and US states, especially those with radically different approaches to taxation (e.g., Japan, Oregon, etc.), if I can find the appropriate information.

Some means of figuring in the proportion of the population that is in each income range would also be interesting.

I’d love feedback and suggestions for this, both regarding the overall approach, as well as the details of each tax system. For specific issues I need help with, see below.


The approach and data here do have limitations. For example, it can only approximate the costs associated with healthcare in the US (go to the US - CA spreadsheet to twiddle the numbers), and doesn’t take into account many special situations and more specialised tax credits and deductions.

Perhaps more importantly, it doesn’t attempt to quantify the quality of the services you get for your tax dollar — arguably, one of the more important considerations, considering how similar many tax regimes turn out to be.

Also, because PPP is a country-wide measure, it doesn’t take into account more specific markets; for example, housing as well as gasoline in the San Francisco Bay area are more expensive than that in Boise, Idaho, but they both have the same PPP.

Furthermore, this spreadsheet focuses primarily on income taxation. Although I plan to add information on capital gains, there are other costs — both tangible, such as property taxes, fees and duties, and intangible, such as government bureaucracy — to be considered when comparing jurisdictions. Note that sales taxes, such as GST and VAT, are already included in PPP.

Preliminary Observations

As I said, this is missing a lot of information and isn’t well-tested yet, but let’s try to take a look and get an idea of what’s going on (any help in correctly interpreting the data would also be appreciated).

Here, we have a two-adult household, both 35 years old, with one person making 65% of the total income in wages and both saving 15% of their wages for retirement.

What do we notice?

It’s good to be well-paid in the UK. This surprised me; many Americans think of England as socialistic, but in fact it’s the least progressive of the bunch, and positively friendly to those with higher incomes. That’s not saying you’ll get as high of a salary in the UK as you would in the US, of course, but if you can manage it, it looks to work out well for you.

Australia is the most progressive of the bunch. Granted, the US and UK’s figures don’t have many low-income benefit schemes built in yet, and Sweden doesn’t have many of the social benefits calculated yet either, but the simplicity of the Australia tax system — even for those not lucky enough to make a good wage — is refreshing.

The US really, really wants the wife to stay home. Compare this graph — where one person stays home and the other works — to the previous one;

Every country except the US taxes per person, rather than per household; as a result, the US effectively encourages one person to stay home. Horrible social and economic policy, all in one package!

The mortgage insurance tax deduction is evil. Have a look at this graph, compared to the first;

What’s the difference? It assumes that the household has PPP$35,000 of mortgage interest a year; America is one of the few countries that allow it as a deduction (the UK stopped in 2000). This effectively encourages people to stay in perpetual debt, and can also be thought of as a renters’ tax.

In fact, I’m beginning to think that, in America, if you don’t toe the line and buy a house in your twenties, stay in perpetual debt for most of your working life, and then make sure you have it paid off before your income declines, you’re pretty seriously economically disadvantaged. Land of the free, indeed.

The big takeaway here is that most countries aren’t that dissimilar. Australia is at most a few percentage points more than America at the higher incomes, and lower across the board when you save more for retirement (which AU allows you to do much more of than the US, BTW; see the second graph on the summary worksheet). Even looking at Sweden in the first graph above — which doesn’t have many benefits factored in yet — the differential in tax rates is dwarfed by that of the services you get in return.

Open Questions

Help with the following specific issues would be very much appreciated;

Once again, IANAE, and for all I know, this may all be based upon quicksand. If you are, I’d like to hear what you think about this, whether I’m reinventing the wheel, etc.


Jesper Joergensen said:


I believe Scandinavian countries have shifted a lot of their taxation from income to sales taxes and other consumption related taxes. I don’t know Sweden that well, but I am Danish and I believe the two systems are pretty similar. In Denmark, sales tax is 25% and it is applied to everything. Vehicle registration is 180% (you pay about 3 times the factory price of a new car). In total, I think that about half of the country’s tax revenue is from income tax and the other half is from sales tax and similar sources.

In general most goods you buy in Denmark are more expensive than in the US. That means that your final measure of “wealth” will depend on how much value you put on material goods vs. e.g. free time.

Regardless, this is an important parameter when you look at income tax vs. the services offered in Calif and Denmark/Sweden. While it’s probably still true that you get less for your tax dollars in Calif, the difference gets exagerated when you only look at income taxes because of the proportion of tax revenue coming from other sources in Denmark (and I believe Sweden is similar to Denmark here, but I could be wrong).

Another interesting thing about this is that sales taxes (particularly when they are applied to everything) are regressive. Thus, Denmark, normally considered to be socialistic with high degree of progessive taxation, actually has a huge chunk of tax revenue coming from a very regressive source. On the other side, this form of taxation encourages saving, so it has it’s good side too.


Monday, July 5 2004 at 10:12 AM

Fredrik Lundh said:

In Sweden, the sales tax is 25% on everything, with exceptions for food (12%), newspapers and books (6%), cultural activities (6%), apartment/house rentals (no tax), and few other things.

As for the pension system, the following link provides an overview of the national system:

Most people also have contract pensions (via their employers), and private pension insurances. Try searching for “pensions” on the Swedish Institute’s homepage ( for more overviews.

Wednesday, July 21 2004 at 5:35 AM

Jorg said:

I’ve run a horticultural business for 35 years before retiring last year. We had 17 corporate taxes levied on us. I believe food may not be taxed directly at it’s final sales point to the consumer, but the company that produces it is surely taxed from the machinery to the land. Wouldn’t you have to figure in tax on food then? Thanks, Jorg.

Thursday, October 28 2004 at 6:04 AM

Jon said:

Wow, interesting analysis. I think youve got two problems though in this initial posting, and they’re really more in analysis probably based on your unfamiliarity with some of these systems (California’s particularly) in actual practice. One is that taxing married couples together allows the gender pay gap (which does exist) to be spread among two people in their taxes, thereby encouraging marriages (and incidentally families) This seems to be a good thing not a bad thing socially and economically. Likewise, the “renters tax” otherwise known as a mortgage deduction encourages home OWNERSHIP…again a good thing. Its really the only way you can really be sure to encourage people to save for a rainy day in terms of tax policy. And the perpetual debt argument is crap, sorry, but it is. You are MUCH better served by having low interest debt in a mortgage than paying out the wazoo THE SAME AMOUNT OR MORE for rental property. Remember, at the end of a mortgage(or at any point when you can’t afford it) you have a home or at least a lot of equity that you can sell for real money, at the end of a lease you have to get out. You seem to forget that there is such a thing as “good debt”

By the way…for a revision you suggest using Orgeon as being very different than California…I’d actually suggest you check on some midwest or southern states, they have real differences …California and Oregon actually i think you’ll find have a lit of similarities compared to those.

Sunday, December 11 2005 at 10:21 AM

Jon said:

By the way…in the UK dont forget the VAT at each level of taxation. It might not be correct in the information if the UK seems to be low.

Sunday, December 11 2005 at 10:27 AM

Creative Commons