TwitterNZ, tax, and transfer pricing

BusinessDesk reports that Twitter is setting up a local affiliate. So we can expect, a year from now, performative outrage about local tax paid. 

When I taught public econ at Vic, we had a short bit on company tax and international considerations. I invited the students to imagine that there were no Google New Zealand. Instead, New Zealand companies would be invited to bid for the right to manage NZ-based advertising sales for Google, getting a cut of the revenue from the adwords auctions. In the limit, where NZ is a competitive marketplace, how much do Kiwi companies bid for that right? The price of the right to be the contracted local affiliate should eliminate any local economic rents. The rents instead accrue to the IP, developed, run and taxed in the United States. 

Under that setup, the local affiliate doesn't pay much tax in NZ. It collects a lot in revenue, which it pays as its license fee to the US. Company tax is based on revenue less cost, and that license fee is part of the cost. It's revenue to Google US, and then gets taxed there.

It's obvious when you run it as separate companies. When instead Google NZ pays Google US to get the advertising product it onsells here, people argue about the transfer pricing arrangements and whether they're set to replicate the numbers that you'd get in the separate-companies kind of scenario, or whether they're set to shift taxable income to places where taxes are low. There are OECD transfer pricing guidelines on this stuff, and I understand that Google works with IRD to make sure that their practice is sound. 

Like, imagine that people were outraged that, while Honda makes a lot of money selling cars here, they write off as expense the cost of buying the cars in Japan and shipping them here, and called for taxing Honda based on gross revenues. Is it that different from what Jonathan Milne says here, from today's Newsroom Pro newsletter?

Facebook doesn't disclose its earnings in New Zealand, but the G7 agreement coincides with Google NZ filing its annual accounts with the Companies Office. Its press statement highlights the $3.3m tax it is paying on a profit of $10.19m. It fails to mention the $517m that Google NZ has paid its US parent company in service fees – an expense that it books against its gross earnings in order to report a laughably small taxable revenue in this company.

Anyway, maybe there's some advantage in Twitter bringing its NZ sales in-house and having local staff. But expect outrage next year when similar amounts of money go back to head office to cover the IP, overall tax paid doesn't much change, but the optics of it do because it'll be a transfer pricing arrangement. 

Update: Pattrick Smellie at BusinessDesk is worth reading. You should subscribe. 

However, the actions of Facebook and Google in NZ also indicate how differently the big tech companies can respond to such pressure. 

It is clear Google NZ is inclined to make nice, choosing in the last two financial years to declare advertising revenue booked by its NZ office paying tax on that. 

It presents an activist face in its government relations and is trying to stay ahead of a government showing increasingly regulatory instincts. It will placate local news media by implementing its Google News Showcase product, which will pay news producers something for republication and is being hailed in Australia as reviving news media companies' fortunes. 

In short, Google is pursuing a self-interested, carefully calibrated exercise in good corporate behaviour to maintain its political and social licence to operate in NZ. 

Facebook apparently doesn’t give a flying one. 

The company has a few employees in NZ, but the only time this reporter has ever clapped eyes on a Facebooker in an official capacity was before the 2017 election when executives from Sydney hosted a breakfast to explain how journalists could do a better job of providing free content for Facebook to monetise. 

At the same time, the company has simply stopped disclosing anything about the scale of its commercial activities here. 

Where Google pursues what it hopes is enlightened self-interest, Facebook’s approach is cynical - and arrogant to the extent that it has been flouting its filing obligations for five years.

Capital Gains Tax Bleg

When I first started writing posts on capital gains taxes three years ago, starting with this post and this one, they were sort of a bleg. I have never understood the rationale for CGT, as the arguments that are usually put seem to involve shifting definitions, or incomplete partial-equilibrium analysis. So I wrote those two posts to explain why I thought the arguments in favour don't add up, hoping that someone could counter with a coherent argument. With a CGT defended within the context of a coherent model, it should be possible to phrase the debate in terms of differences in either values or empirical beliefs about the economy. Three years on, I have seen a lot of public discussion of capital gains taxes, but still don't understand what is the model from which proponents draw their conclusions.

But now I have a different bleg. I would like to know how actual CGTs that have been implemented elsewhere (or the ones proposed by opposition parties in New Zealand) would deal with a particular issue. This issue is easiest explained with a series of examples:

1. Imagine that you are a householder with a portfolio of $2,000 of shares in a single company that is earning a 5% rate of return on its capital. At the start of a new year, you decide that you want to save some more, so you buy $100 in a different company using money you have earned in the previous year but not spent. You now have a portfolio of $2,100. This increase in the value of your portfolio would not be classified as a capital gain. It simply represents increased savings.
2. Now imagine that instead of putting the $100 into a different company, you bought $100 of freshly issued shares in the same company that you already have a shareholding of $2,000.  That still doesn't count as a capital gain, right? The company uses the money from its new share issue to buy some capital equipment, which will also earn a 5% rate of return, but what they do with the money is irrelevant.
3. Now, imagine that in the previous year, you were paid a dividend of $100 by the company in which you own $2,000 of shares, and you bought $100 of freshly issued shared by the same company. Again, this makes no difference. The fact that your purchase of new shares was in exactly the same amount as your dividend payment, is irrelevant; the additional $100 was paid for out of your total income and was available for buying shares because of a choice not to use it on consumption. 
4. Now make one more change. Instead of paying out $100 in dividends and then issuing $100 of new shares, the company simply retains the profit, pays a dividend of $0, and uses the money to pay for the new capital, as above. The company has increased its ownership of capital equipment by 5%, and so the value of the existing shares will increase by 5%. So now our shareholder sees that his shareholding portfolio has increased from $2,000 to $2,100, just as in all the previous cases, except in this case he hasn't bought any new shares; he has seen what looks like a capital gain of 5%. Except, it is not really a capital gain; the reinvestment of profits by the company instead of paying out a dividend is a form of saving that is imposed on its shareholders.  

In the absence of taxes, the only difference between example 3 and 4 is in the default position. In 3, the shareholder receives the dividend and needs to make a decision to buy new shares to turn that dividend into saving. In 4, the default is that the dividend is saved, and the shareholder would need to sell $100 of shares to convert that saving into consumption.

But what if we add capital gains taxation. Wouldn't a CGT induce a difference between example 3 and example 4, adding additional taxation in the latter but not the former. And wouldn't this induce a distortion in which the tax system created an incentive for firms to pay out all their profits as dividends and then raise new capital rather than retaining profits for investment? What I would like to know is how to other countries deal with this distortion in their CGTs (if at all), and how would the opposition parties in New Zealand plan to deal with it.

It is also interesting to note that currently in New Zealand, there is a slight tax distortion that favours retained income over dividend payout (as the corporate income tax rate is lower than the top rate of income tax). At the time the rate was lowered from .33 to .30 in 2008, the then Labour government said this was a deliberate distortion as a kind of nudge to encourage retained earnings and hence make saving the default. Does Labour now think that we should be changing the nudge to consumption by moving the tax advantage the other way?