Friday, March 18, 2016

Hawaiian Taxes, Vertical Integration, and a Research Project

Not a Sales Tax

I recently spent a few days in Hawaii, teaching some classes and giving a talk at the Iolani school. In the process, I learned something interesting about the Hawaiian tax system. A major source of state income is the general excise tax, a tax, in most cases of 4% (4.5% on Oahu), on the gross revenue of firms. At first glance that looks like a sales tax, but it has rather different implications.

Consider the taxation of bread. A farmer grows wheat, sells it to a miller, pays 4% tax on the money he receives. The miller converts the wheat into flour, sells it to the baker, pays 4% tax on the money he receives. The baker sells the bread to customers, pays 4% tax on what they pay him. The wheat has been taxed three times, for a total tax of (because of compounding–taxes on taxes) 12.5%. The miller's contribution, the value added to the wheat by milling it, has been taxed twice. The contribution of the baker has been taxed once. Depending on the relative amounts of value added at each stage, the total tax on the bread is between 4% and 12.5%.

Suppose, however, that the bakery is a vertically integrated firm, owns its own farm and mill. That makes the transfer of wheat from farm to mill and of flour from mill to bakery internal to the firm. Only the final stage, the sale of the bread, produces revenue to be taxed. Now the total tax on the bread is only four percent.

Which means that one effect of the excise tax is an artificial incentive for vertical integration, making firms in Hawaii larger than they would otherwise be.

A Theory of Exploitative Taxation

While on the subject of taxation in Hawaii, an interesting conjecture occurred to me, which I thought someone else might be interested in pursuing. Consider two adjacent states that are equally attractive in terms of climate, culture, and the like. They compete for taxpayers. If one of them has higher taxes and does not use them to produce a correspondingly better level of services, it will lose citizens, and taxes, to the other.

Consider the same story, with one change—one of the states is, for reasons independent of taxes and government services,  a pleasanter place to live in than the other. It can collect taxes and use them in ways that produce no benefit to the taxpayers—to buy the votes of state employees, say, with generous wages and pension benefits. As long as the cost of the higher taxes is lower than the value to taxpayers of the state's natural advantages, they have no incentive to move. 

Hawaii has unusually high state taxes per capita. It is also a very pleasant place to live—comfortable climate, lots of beaches. 

To turn my conjecture into a research project, you need data of two sorts. One is some measure of the level of exploitative taxation in each state—tax collected minus value of services to the marginal taxpayer/resident. The other is some measure of each state's natural advantages. If the theory—think of it as an elaboration of the Tiebout hypothesis—is correct, the two should be positively correlated.


Anonymous said...

Sorry for not understanding; the sales tax - it is not deductible for the (business) buyer? I.e like the VAT of European nations?

David Friedman said...

As I understand it, it's a gross receipts tax not a VAT or a sales tax.

Joey said...

David, how are you getting 12.5%?

Seems like the most it can be is 12%:

Farmer sells wheat for $100, pays $4 in tax
Miller adds $4 of value, sells flour for $100, pays $4 in tax
Baker adds $4 of value, sells bread for $100, pays $4 in tax.

Final sale price is $100, total tax paid = 12%. Looks like a global max.

If in each step no value were added, you'd have an effective rate of 11.5% because of compounding downwards.

LH said...

Three interesting data points would be California, Texas, and Florida - all large states in population and geography that have at least some favorable climate. California and Texas are large enough that perhaps a higher level of granularity would be needed in segmenting the geography. California does have high taxes, while Texas and Florida don't have a personal income tax (I'm not quite sure how the other taxes add up). California would probably support the theory, while Texas and Florida might not. I think the theory is probably correct but the effect might not be large enough to prove with state-level data.

Hawaii has, arguably, a higher barrier to exit than the other states as well as an even more favorable climate, so the effect should be particularly pronounced there, IMO.

There are probably some good county-level weather data sources available these days, and those could should provide a sufficient level of climate/weather granularity while comparing against state taxes. I'm tempted to try such an analysis, although I wonder if it would be accurate without factoring in municipal taxes.

I was wondering about the 12.5% calculation, too. Like Joseph, I got about 11.5%.

David Friedman said...

Farmer requires $100 net revenue to pay for producing wheat. Sells the wheat for $104.17 from which he pays a tax of $4.17, which is .04x$104.17.

Miller requires $104.17 to cover the cost of the wheat, aside from any other costs. Sells the flour for a price that includes that and the 4% tax on that: 108.51. Repeat for one more round. I slightly underestimated the total because I forgot that the first round is paying tax on a price that includes the tax.

My initial calculation assumed that the tax was on revenue net of tax--that the buyer paid the price plus the tax and the seller was only taxed on the price. Doing it that way, the farmer gets $100 for his wheat, the miller pays $104 (price plus tax). The miller then sells the flour to the baker for a price that includes $104 for the wheat, so the tax on that is not $4 but $4.16, including the tax on the part of the price that came from the previous tax. When the baker sells the bread, the part of the price coming from the wheat is $108.16, so the tax is $4.33. Add up the taxes and you get $12.49.

Either way, you have to allow for compounding--for the fact that the tax pushes up the price at each level, which pushes up the price at the next level, which increases the tax at that level.

I was originally going to make it "about 12%" but I figured someone would call me on the compounding so I should include it.

David Friedman said...

I'm wondering if you are doing your calculation on the tax as a fraction of what the consumer finally pays, where I'm doing it as a fraction of what he would pay if there were no taxes at all.

I believe mine is the usual way of defining a sales tax. If the sales tax is 10% on a $100 item, that means the consumer is paying $10 of tax for a total payment of $110, and $10 is not ten percent of $110.

Joey said...

I'm wondering if you are doing your calculation on the tax as a fraction of what the consumer finally pays, where I'm doing it as a fraction of what he would pay if there were no taxes at all.

Yes, I was. And I didn't assume that the prior tax on the inputs would count as a tax deduction, but they probably should.

So using my numbers (which roughly maximize the final rate of taxation in the same way that yours do), and calculating as a percent of what the consumer would pay if no tax including deductions, we get your 12.5%:

(4 + .96*4 + .96^2*4)/(100-(4+.96*4)) = 12.5 about.

The amount for the final price that the consumer would have paid is not $92 but a tiny bit more now because the tax deduction meant that the taxes were a bit smaller.

Harry Messenheimer said...

David, I did a study of link between economic freedom and prosperity a few years ago along the lines you suggest. It's more a study of overall government exploitation, not just taxes. I compared interstate counties along New Mexico's border. Adjacent counties along the border appeared to be similarly "attractive in terms of climate, culture, and the like." Study is on Rio Grande Foundation's site:

Also, I don't think Hawaii's gross receipts tax is as bad as you described (although overall tax/government exploitation here is pretty awful). "Wholesaling, Manufacturing, Producing, Wholesale Services, and Use Tax on Imports For Resale" here are taxed at the rate of 0.5 percent rather than 4 percent.



David Friedman said...


I did say "in most cases." As best I could tell from trying to make sense of the rules, a wholesaler selling finished products to a retailer only pays .5%. But a firm selling inputs to another firm pays the full 4%. As in my example.

Am I mistaken? Do the farmer and the miller in my example only pay .5%?

Jonathan said...

Looking at the range of tax rates in the world (, I don't see an obvious correlation between pleasantness and high taxation.

Roughly speaking, the highest tax burdens are in northern Europe and the lowest in oil-producing countries.

Unknown said...

"Wholesalers get a special GET rate of .5%.

What’s a wholesaler? Someone who sells goods in bulk to other businesses to sell for retail. "

It's is indeed possible that both the farmer and the miller count as wholesalers, but if the farmer has a PYO (Pick Your Own) sideline maybe he has to register twice ?

Same if the miller sells to a "end user", oatmeal to a dude ranch ?

VAT was invented to get rid of similar turnover taxes.

On the taxation by state conundrum, I understand that US states differ, some mainly use income taxes, some sales taxes and some property taxes. So irrespective of the general tax level they will have a differential impact anyway.

David Friedman said...


Neither the farmer nor the miller in my story is selling goods in bulk to other businesses to sell for retail. The baker isn't reselling flour, he's selling bread.

How the Hawaiian tax authorities interpret the rules I don't know.

Haris H. said...

Everyone hates links but I wrote about exploitative taxation this in another context:

Not an empirical study, but essentially a restatement of the theory. Certainly Texas and California provide two very large data points consistent with the theory.

Haris H. said...


I don't think a global comparison is very informative because it's almost impossible to control for other factors, specifically culture & immigration restrictions. Swedes speak Swedish and enjoy Swedish culture; they won't move to lower-taxed Azerbaijan on those grounds, and Saudi Arabia might not even let them in. This significantly reduces the pressure on governments to be less exploitative.

You'd also have to account for political variables - Saudi Arabia has low taxes because it has oil money and its people would probably rise up if they were taxed highly, too. Sweden, meanwhile, is the opposite, with a responsive democratic government. Impossible to compare the two.

Unknown said...

Ah yes

poking about on the Hawaii tax site it appears that the
0.50% rate is for Wholesaling, Manufacturing, Producing, Wholesale Services, and Use Tax on Imports For Resale

Jonathan said...

Haris H: OK, then how about limiting the comparison to the countries of the European Union? I've lived and worked in six of these countries; it's easy to move between them, they offer relatively compatible cultures and lifestyles, their languages are more or less related to each other, and many of their inhabitants are multilingual.

To take your example of Sweden, many Swedes speak English well, and they quite commonly speak another European language too, such as German or French.

I now live in Spain, a pleasant and relatively warm country that receives quite a lot of immigrants from other EU countries (and from north-west Africa). But I think its taxes are quite moderate by EU standards.

Tibor said...

Jonathan: Culturally, the individual European countries may not be that different (although I've had the misfortune to have to listen to a lecturer at a mandatory 'diversity competence workshop' who was a German who moved to Austria and basically kept talking about how hard it was for her to adjust to the different culture...but anyway), but the climate is. I probably would not feel happy about living in Spain (or Sweden) just because of that. Sweden would be particularly bad, because of the extremely short days during winter (and subsequently unnecessarily long days during summer), in Spain I would miss the forests. I guess that the Spanish would find Swedish winter even worse, Swedes would probably miss the snow (although I was told by a Spanish colleague that there are places in Spain where it does indeed snow sometimes). Then again, you are right that the difference is far less extreme in terms of both culture and climate than say UK vs. Saudi Arabia.

Still, I think that although each factor may play just a little role, in sum they are quite strong. Europeans are generally quite unwilling to move very far. I am not not talking about people with higher education who can all speak English (regardless where they come from) and associate mostly with people who also speak English, so the language barrier is not a big deal for them, who also tend to be more "international" in their habits in general and not so attached to their place of birth. I am talking about the average Joes/Joses/Josefs/Giuseppes who usually prefer to stay in the same town they were born. I think this is probably quite different in the US and partly something that can cause Spain to have an unemployment rate of 20% while Germany has 4.3% and both countries are a part of the Schengen zone (true, one might need to speak German, especially for the less qualified positions). At the same time, in the US it seems to be considered nothing extraordinary for people to move farther than one even physically could within Europe just to get a better job.

As for languages - I don't think they are that similar. English is probably the most "connected" European language, being similar to both Romance and Germanic languages. But there are not many connections between Spanish and German or between Czech and Italian. Germanic languages are similar to each other, Romance and Slavic are too and some languages from one group have connections to some from a different, for instance for me, the most fun part of learning German is discovering the words or expressions which I thought were Czech. But generally you do not have too much overlap between these three groups (English is a notable exception) and to make things worse, you also have a couple of odd languages like the Finno-Ugric group (which are not even Indo-European), Greek, Baltic languages (other than Estonian which belongs to the Finno-Ugric group), not to mention Basque (but of course, there are probably Basque speakers who also do not speak Spanish). Sure, they all have very similar grammar structures (although the Finno-Ugric with their system of dozens of grammatical cases which completely replace prepositions not so much) and all use the same alphabet except for Greek and Bulgarian (in the EU).

Tibor said...

I think Latin American countries would make a better comparison. I don't know how strict the immigration restrictions are there, but the whole continent speaks two very similar languages (ok, Dutch in Suriname and English in Guyana, but those are a tiny part of Latin America), the culture is even more similar than that between European countries and even the climate is for the most part similar-ish except for the southern parts of Chile and Argentina.

Another good case might be Switzerland. The cantons are about as independent as the US states and set their own taxes. They are also tiny so that the transaction costs of moving from one to the other are much lower than in the US (visiting friends from a different canton is a question of a 1 hour train ride as opposed to a 4 hour flight). They do not all speak the same language (and indeed there seems to be a difference between taxation in the German-speaking and French-speaking cantons, the French-speaking tend to have higher taxes...there is only a single exclusively Italian speaking one and none where Romansch is the only official language), but even if you restrict yourself to the German-speaking and partly German-speaking cantons, you still have 21 datapoints.

Jonathan said...

Tibor: Yes, your suggestions of South America and Switzerland are good examples.

I've found it easy and comfortable to move around in Europe; admittedly, it helps that my first language is English, which many people can speak.

I lived in Sweden for four and a half years without difficulty. The buildings are all well heated, so you feel the winter cold only when you go outside, and then you just have to wear suitable clothing. The summer is similar to English summer, except for the longer days.

There are forests in parts of Spain, and snow falls every year in much of northern Spain; though only rarely where I live now (near Barcelona). The climate varies a lot from one part of Spain to another.

Patrick R. Sullivan said...

It definitely snows in Spain. I've seen it not far from the Mediterranean in the Sierra Nevade mountain range. There are plenty of ski resorts in the country.

Washington and Oregon are a natural experiment in the Pacific Northwest. Washington has a sales tax, but no income tax, vice versa for Oregon. Climate is about the same.

Tibor said...


Sure, for the speakers of English, even if it is not their mother tongue, moving around Europe (or rather the EU) is easy. Although I still think that if you don't learn the local language, you won't end up feeling very comfortable. Or perhaps it depends on your tastes. I've always found quite strange the kind of people who move to a foreign country only to hang out with other "expats" from their own. But in any case, it is easier than if nobody spoke English. Then again, in Germany, at least where I am (and it is a university town, although quite small one), you will have a hard time with just English in some situations. Obviously, at the university, everyone speaks English very well, but you cannot expect that from a hairdresser for example or even most clerks. When I was setting up health insurance - my German was far worse then than it is today - there was a single woman in the whole office who would speak English and when I was setting up my bank account, there was nobody, so I was signing something I could barely understand.

I guess you get used to the weather, but even if you're fine with good clothing, the experience of the outside world is different when it is -20°C outside. For one, you usually don't want to walk around in a ski mask, so your face is freezing even if you're wearing good clothes otherwise. Actually, I think the lowest temperature I've ever experienced was -20°C or so for a few days and I just tried to go outside as little as possible...I cannot imagine how people cope with the -30°C winters in Canada or northern Scandinavia where these temperatures seem to be quite ordinary.

My view of Spain is probably a bit distorted. I basically imagine La Mancha and assume that all of Spain looks like that. I've only ever been to the Baleares and Lisbon (which is obviously not a part of Spain but close enough).

Jonathan said...

Tibor: The advantage of English as a first language is that I've usually been able to WORK in English in any country. In everyday life outside work, the people don't usually want to speak English (even if they can), so I've made some attempt to pick up German, Italian, Swedish, French, and Spanish, in that order. I'm not good at acquiring languages, but I acquired at least a little of each language.

I don't try to hang out with other English people, so I've spent many hours in bars and restaurants with people I could hardly understand. And then I married a Spanish lady. :)

I spent four and a half years in Stockholm, and the temperature never went as low as minus 20. Just one night, it was minus 19, and in the morning my car refused to start, so I had to walk to work: 45 minutes through the snow. During the winter, the temperature is more commonly close to zero, although it could be minus 10 sometimes.

Spain is a very varied country. Some parts are dry desert, but the north-west (Atlantic coast) is cool and damp. Here in Catalonia, the Costa Brava is green, and there are forests inland, but I live in a semi-arid area. There are plants and trees, but they live with some difficulty, because there isn't enough rain. The countryside is covered with fields of grapevines, for making wine. There are also olive and almond trees.

Tibor said...

Jonathan: I think that with a sufficient level of English, (let's say at least C1), you can work in English anywhere even if it is not your first language. But there is a caveat - only the higher education positions allow this, for most others (actually probably save for the very low-skilled positions like cleaning), you typically have to work in the local language There is a notable exception - language teaching. And since English is taught everywhere, even someone who lacks any special skills but who is an English native speaker can find a job pretty much anywhere without the knowledge of the local language. German or French native speakers will still have it easier in this respect than the Dutch for example, but it will still be harder for them than for the English natives.

I used to have the opposite problem - when people saw that my German (which has meanwhile improved significantly) was not very good, they started speaking in English automatically (most of the people I meet come from the university and so they all speak good English). That way it was really hard for me to improve and I had to take a course first to get to a level where people would not switch to English after two sentences when talking to me. But this might be a specific feature of a student town. Then again, my former English teacher who lives in Prague told me that it is really hard for him to improve his Czech, because once people find out he's American, they always want to practice their English with him. On the other hand, I play cajón with a group of Peruvians here and they often talk in Spanish at the rehearsal even when I am there...which might be considered a little bit rude, but it was a good motivation for me to start learning the language and since Spanish is quite easy, I can now (after some 7 months of playing with Duolingo from time to time) usually follow at least 1/3 or 1/2 of the conversation, more if someone talks to me directly (although that they usually do in German or English). Similarly, people told me that in France, people are not very keen on talking in English (even when they can), so that also forces you to learn the language.

I guess I should visit Spain (the mainland) myself...also to practice peninsular Spanish:) I have absolutely zero knowledge of the "vois" informal plural, since it is not used in Latin America at all (they only use ustedes, even with friends).

Since you live in Spain, one question you might be able to answer - How much does the 20% unemployment rate correspond to reality? The number seems extremely high to me, I cannot imagine a European welfare state coping with such a high unemployment rate for very long without ending up like Greece. I was told that in practice the number is much lower since a lot of people work illegally (thus avoiding all the rules that make employing people in Spain extremely expensive). Do you have any estimate for what the "real" unemployment rate is?

Jonathan said...

Tibor: I found in Italy that Italians are not usually keen to speak English, even when they can; so I learned enough Italian to use it at work.

In Sweden it was quite different. If I tried to speak Swedish, the Swedes would shift to English; and sometimes they even seemed offended that I tried to speak to them in Swedish. I was insulting them by implying that they might not be able to understand English...

Then, after I'd lived in Stockholm for about three years, they started asking me why I hadn't learned Swedish yet. (Of course, because I never got a chance to practise...).

Here in Spain, I find that "usted" and "ustedes" are rarely used. In some circumstances yes, but not among friends or at work.

I can't tell you much about Spanish unemployment. I have work, so have most of the people I know. I suppose that many unemployed people are living with employed people; for example, my wife is unemployed, although she has worked at times in the past.

Gennie said...

Keep sharing more informative posts like that,