GST at the border

Bronwyn Howell kindly walked me through what seems the least hassle-ridden way of collecting GST at the border (other than Seamus's proposal): offload it onto the domestic end of the good's shipment. The mechanism, if I understand it properly, would work as follows, with all confusions being mine.

When you order something online from a foreign shipper, they have to get it to you. Whoever they're using for the international leg has arrangements with a domestic shipper to get it to your door. And there aren't that many courier companies running the domestic side of things here, plus NZ Post.

On the local delivery agent getting a heads-up that a package is on its way, it would also have to get a heads-up on the goods' stated value. The delivery agent would then be liable for GST on the shipped item.

That agent would then contact the recipient of the good seeking payment - when you order something online, you're giving them an email address anyway. When you get the email, you'd go to another website to pay the tax so that the shipment could get to you immediately on landing in NZ. If you don't get around to doing it before the product gets here, or if it gets lost in your spam filter, you'd have to pay the courier when the item gets to you or pick it up from the courier office and pay there. In a competitive shipping market, we'd expect the shippers to figure out easy ways to facilitate payment, like keeping your credit card details on file (if you agree) so that you can be automatically billed for any future GST charges without having to get an email.

So, in a best-case world, if you'd already paid GST once through that shipper and hadn't changed credit cards since, you'd just get an email noting that the shipping company was going to charge your card GST on a shipment that's coming through, with opportunity for you to object if something didn't make sense.

That would be pretty hassle-free, after the initial hassle of having to set up accounts with the different shippers.

I note, though, that UPS in Canada always managed to charge us about $40 for the service of paying the duties and tax on our behalf at the border. Note that these fees are over and above any actual taxes collected. One hopes that New Zealand's domestic shipping industry is sufficiently competitive that that wouldn't happen here, but I doubt that the shippers would agree to become tax collectors for free either.

I also wonder whether the process might provide a mechanism for griefing shippers and others, if anyone were so inclined: ship a thousand envelopes each with a very high customs valuation for a photocopied picture inside to a bunch of unwitting recipients. The automated systems would either charge them automatically for the GST on something they'd never ordered, or would trigger collection and confusions, or would require the shipper to send back the envelopes while explaining to Customs that no GST was collected on the items. Maybe there are no griefers out there who'd try it though.

We're then weighing up the transaction-deterring hassle costs (albeit perhaps smallish in this case), combined with the extra cost imposed on consumers by turning the shipping companies into tax agents, against the allocative efficiency gains from removing a tax distortion and the prevention of the erosion of a part of the tax base. And, at the same time floating around in the background, the expectation that if the system does wind up doing too much to deter consumer-led parallel importation, domestic retail prices would likely go up proportionately.

The system seems worth IRD's investigation. But assessing the benefits of it really shouldn't begin with the numbers in the ISCR report commissioned by BooksellersNZ. The report has a lot of good stuff in it - it's where I saw the scheme noted above, and why I got in touch with Bronwyn.

But the report does have a few ...issues.

1. The report gives, as one option, requiring foreign sellers to register with IRD. While I can buy that a lot of large online retailers would find it worthwhile to sign up to the kind of multilateral system they describe, I doubt that that is also true for lots of the smaller online US retailers. For many of them, any international shipping already seems a hassle: that’s one reason YouShop was set up, right? To forward on packages from US retailers who don’t want the hassle of international shipment?


2. The report suggests that, under the system where foreign firms would register with IRD, local firms would have to pay taxes to the US on shipment there through some OECD coordinating mechanism. But wouldn’t that kind of system, for shipping to the US, be next to impossible absent the US sorting out its mess of local sales taxes? I mean, they’ve not yet been able to sort things out for internet sales taxes within the US; there was some talk a couple years ago about having a set of states that agree to a reasonably common base also agree to collect sales taxes for each other, but I don’t think it’s gone anywhere, has it? There are thousands of local taxing jurisdictions in the US with really divergent rules over, for example, what precisely counts as an ice-cream sandwich (and what doesn’t) for sales tax purposes.


3. They suggest a few methods for evaluating whether shifting the regime would be a good idea and argue that a social welfare maximisation approach is strongly preferable to a government-centric approach in calculating the appropriate de minimus value or in setting other parts of the collection regime. I agree, but have a couple of worries on this front.


1. Absent a system that is able to collect these fees pretty seamlessly, from the consumer’s perspective, we need to watch for spots where we might impose fixed costs on purchases from abroad that do not obtain for shopping domestically. For example, it seems likely to be pretty distortionary and welfare-reducing if customers are deterred from buying from abroad because of a few days’ processing lag at the border, or because of a requirement to go pick up the item at a NZ Post Office across town and there pay the duties, or an additional step when shopping requiring you to go buy a customs stamp from a Customs website.


2. There's no note of that easy ability to parallel import from abroad can serve as substantial constraint on price-setting in domestic retail markets. That magnifies that harm that could be done if we unduly deter customers from using foreign shopping options due to expected hassle-costs. Sure, it's a second-best-worlds consideration, but we'd need some accounting for it in a social-welfare-maximisation setup.
4. I’m really not sure that the elasticities they're using from Einav et al, for example, would really apply here - they're there used to estimate how much change there would be to shopping patters with the 15% GST being applied on international low-value purchases based on tax elasticities across US states. Shipment from different US states, from the point of view of the consumer – there’s really not much difference other than price. Whether you buy in-state or from out-of-state, you’ll have whatever you ordered in 2-5 days. Here, that’s not quite so: domestic shopping is far more immediate than shipping in from abroad. Since they’re not nearly as close of substitutes for one another, the price elasticity of demand between foreign and domestic should be lower, right? Further, because the NZ market is much thinner, there are plenty of products you just can’t get here: again, this lowers the expected price elasticity and means that you might be overestimating the effects of the de minimus thresholds.


1.  On this one, I’m really willing to put money on it with any of the authors. If a seamless collection of 15% GST at the border is implemented, I am willing to bet that the decline in demand for Book Depository in New Zealand is less than 20% (they predict 45-60%). It’s range of products and ease of getting the books that’s driving demand for Book Depository, as well as price differences well in excess of 15%; an extra 15% charge is trivial – or at least that’s my bet. But only conditional on shipment’s being seamless. I’m more than happy to bet that you could kill demand for Book Depository by making it a hassle to receive shipment.


5. I am curious that they're counting as benefit the jobs that would be created in New Zealand subsequent to deterring imports by imposing substantial delays on importation. Mightn’t we need some modelling of whether those workers would be drawn from other actually productive sectors? It looks to me like we’d be imposing large and real costs here to effect a transfer from domestic consumers and foreign retailers to domestic retailers. Some accounting for the elasticity of domestic prices with respect to degree of foreign competition also seems likely to be relevant.


6. I was interested in their choice of counterfactual at the end where they invite the reader to imagine the benefits of a $5k tax-free threshold if only we could lower the de minimus threshold. Do they believe that this is the relevant counterfactual? Really?

In particular, I was very surprised to see this line in a report that had either Bronwyn's or Norm Gemmell's name on it:

“This enhanced demand at domestic retailers not only results in increased producer surpluses, but firm growth leading to increased employment. Jobs created by a reduction in the de minimis threshold increase the economy’s productive capacity, clearly benefiting the rest of the country. Enhanced domestic employment results in increased consumer spending, and through the multiplier effect, enhanced growth throughout the economy.”

Emphasis added. Jeez.