Saturday, 27 May 2017

A better case for tipping

If New Zealand were to move towards a tipping norm, here's an entrepreneurial idea.

Set up a restaurant. The wait staff are all volunteers. They still have to pass through normal recruitment practices, but they're volunteers. List on the menu prices that there is no charge at all included in the prices for service: none. The wait staff receive no pay except that which is provided by diners as a gift, and list some suggested gratuity levels that would provide, if every diner paid those amounts, various wage levels from $15/hr to $30/hr.

Currently menu prices include not only the cost of wait staff but also the GST that applies on the service provided by the wait staff. My restaurant only charges GST on the non-waitstaff costs.

And restaurants normally have to compete for staff based on salary, but the gifts provided to my staff would all be before-tax. And since nobody's tracking how much the wait staff receive in donations, they'd be on their own recognisance when it came to things like income tax and ACC levies. Really, the tips are a gift, right? There's no gift duty in New Zealand (though I'm pretty sure IRD doesn't consider tips to be gifts).

I might change my mind about this whole tipping thing.

Friday, 26 May 2017

Ending appeasement

Blackland's produced the best summary of the demise of the Wellington 7s that I've yet seen. Bottom line: the only outcome that would appease the critics was the death of the Sevens, so attempting appeasement was not a great strategy.

From their summary:
The organisers acquiesced to, and collaborated with, the “moral panic” (an imbalanced, hyper-sensitive reaction to non-extraordinary normal events) among the City establishment, including the Council and Police.

It was driven by non-attendees who disliked the behaviour, and campaigning media, but gathered strength when it was allied with anti-alcohol and social disorder sentiments. This gave moral imperative and opportunity which meant establishment and elite figures felt obliged to agree that “something must be done” to change the behaviour.

The changes made by the Sevens organisers and the City establishment directly affected the “party” atmosphere. This experience also relied on large crowds – the joint and mutually reinforcing experience. When attendance began to decline in response to the changes, it quickly gathered pace. Each fall affected the experience, which deterred future attendance.

The lesson is that acquiescence to value signalling of noisy people on contentious subjects can disadvantage those most important to your organisation or event; customers, staff or shareholders.

It is instructive that the Police finally praised crowd behaviour and declared themselves satisfied over the 2017 event, when effectively no one turned up and was a fiscal disaster.

The Police and the City establishment had killed a “golden goose” event enjoyed by tens of thousands of everyday New Zealanders, and the Sevens organisers collaborated.  
They conclude:
Conceding that there was a behaviour and alcohol problem tacitly accepted the need for curbs on these factors. That meant conspiring with the moral panic to suppress factors key to the event’s success. When they did, the event lost popularity.

Because the event was an alcohol-imbibing outrageous party, these factors would be a feature. If you remove these features, you remove the Party. You remove the party, you remove the event.

A common response recommended by public relations “experts” is to concede some ground – accept criticism and modify.

This was the first route tried by the organisers. It didn’t work. It accepted that alcohol consumption and behaviour was out of the ordinary and therefore a problem.

This emboldened and legitimised critics. Without the organisers’ backing, no one was standing up for the event’s punters.

Without any defence of the relatively innocuous and common standard of drinking and behaviour, the complaints were not challenged and moderated.

The next route was total collaboration with critics. The organisers introduced rules and components to the event pandered to the idea that it was possible to create a new product and attract new customers.

The critics had no skin in the event. The success or failure of the event would not be their responsibility, and would have no direct impact on them. Their objectives were different. They were motivated by factors such as changing alcohol consumption or attitudes, undermining rugby, signalling virtue to peers, or gaining political advantage or media airtime.
Read the whole thing, as well as their excellent recommendations that follow on from the conclusion.

HT: Stephen Franks.


Product labelling

All we wanted was truth in labelling, backed by tough enforceable regulatory standards, right? Sounds good?

Here's where that path goes:
The fight over the US government’s definitions for certain foods has flared up again. It’s no longer just a fight for milk farmers, who’ve grown increasingly angry about plant-based food companies (think soy, almond, and cashews) calling their liquid products “milk.”

For the first time, vegetables are being roped into the debate—all because of the arrival and popularization of “cauliflower rice.”

“Only rice is rice, and calling ‘riced vegetables’ ‘rice,’ is misleading and confusing to consumers,” Betsy Ward, president of industry lobby USA Rice, said in a statement earlier this month. “We may be asking the Food and Drug Administration and other regulatory agencies to look at this.” Ward added that Scott Gottlieb, the new Trump-appointed FDA commissioner, could use his power to enforce the agency’s existing definitions for food, the so-called “standards of identity.”
The better system: a consumer guarantees act that guards against fraud, and a court system prepared to tell a plaintiff that he's an idiot for suing a riced vegetable producer for having a product that doesn't contain right (and award costs against him).

Thursday, 25 May 2017

New EMTRs!

We have the new Effective Marginal Tax Schedules. Really these things should be published along with the budget documents.

The excellent Patrick Nolan will be presenting some work at the NZEA conference explaining New Zealand's EMTRs. They're kinda high in some income ranges.

I asked Patrick whether he might update one of the main scenarios (they all vary depending on number and ages of kids) to account for the newly announced income tax thresholds and WFF abatement rates. And he kindly provided! Thanks Patrick!

This scenario tracks the EMTR by weekly hours of work for a single-income family with an earner on $20/hour and with one child under the age of 16. I'll quote from the scenario description - note that the description is from the pre-budget version so make your own adjustments.

Again, the $14,000 threshold increases to $22,000; the $48,000 threshold increases to $52,000, the family tax credit increases by $9 per week and abates at a higher rate starting from a lower level of income. The accommodation supplement is not here included.

UPDATE: Updated 16.30 with a correction to the image from Patrick.

And here's Patrick (again: this is pre-changes):
Income taxes and ACC levies are levied from the first dollar of income. The starting tax rate is 10.5% and ACC levy is 1.4%, thus giving an initial EMTR of 11.9%.
Once gross earnings increase to $80 per week the main benefit starts abating. The net benefit abates at a rate of 70% against increases in gross non-benefit incomes. Income-tested benefits are not exempt income under section CW33 1(a) of the Income Tax Act 2007. Thus, as both the level of the benefit and total gross taxable income are above the threshold for the 17.5% tax rate ($14,000 per annum), the gross benefit abatement is 84.8% (given by 0.7 / (1 - 0.175)). The increase in gross income from a dollar earnings is thus 15.2 cents (given by 1 - 0.848), giving an increase in net incomes of 10.1 cents (given by 0.152 * (1 - 0.175) - 0.014), meaning an EMTR of 88.9% (given by 1 - 0.101).
The EMTR remains at this level until the gross benefit falls below $14,000, at which point the tax rate used to gross up the main benefit falls to 10.5%. Consequently the gross benefit abatement is 78.2% (given by 0.7 / (1 - 0.105)). The increase in gross income from a dollar earnings is thus 21.8 cents (given by 1 - 0.782), giving an increase in net incomes of 16.6 cents (given by 0.218 * (1 - 0.175) - 0.014), meaning an EMTR of 83.4% (given by 1 - 0.166).
 The EMTR and remains at this level until the person works 20 hours a week and thus qualifies for in-work assistance. The Minimum Family Tax Credit provides a guaranteed net family income, and so net income above the level of the guaranteed net income ($23,764) abates at 100% until the credit is fully exhausted. With the ACC levy this leads to an EMTR of 101.4%.
 Once the family has a net income before Working for Families of $23,764 the Minimum Family Tax Credit is fully abated, and the EMTR falls to 18.9%, reflecting the 17.5% income tax rate and 1.4% ACC levy.
Once gross income reaches $36,350 the rest of the Working for Families programmes start abating at a rate of 22.5%. The order of abatement is the Family Tax Credit and then the Working Tax Credit. (Note the Independent Earner Tax Credit and Parental Tax Credit are not included in this model.) This gives an EMTR of 41.4% (made up of 22.5% Working for Families Abatement, 17.5% income tax, and 1.4% ACC levy).
At $70,000 gross income the income tax rate increases to 33%, so the EMTR increases to 56.9%, and once the Working for Families entitlement is fully abated the EMTR falls to 34.4%, which is a combination of the top personal tax rate and the ACC levy. The EMTR remains at this rate until $122,063, at which point the ACC levy is no longer charged, and the EMTR is 33%.

Thanks again!

Previously: Bringing sexy back - EMTR style

Tax cuts require fiscal discipline

It's budget day here in New Zealand. Much has been pre-announced; Chris Keall has the summary list over at NBR. But as Rob Hosking points out, there's a pretty big remaining surplus that could yet give us some surprises. 

I'd expect most of those surprises to be saved for later election promises, but it would be nicer if they were laid out in the budget.

When it comes to tax cuts, though, we need to be careful. The surpluses look fine for the next few years, but current tax cuts would need to be reversed in a decade's time if we've neither sorted out the costs of an aging population (NZ Super, health), nor increased productivity and economic growth. And temporary tax cuts do less good than permanent ones.

I cover it off over at the Spinoff.
The case for more substantial tax cuts is sound, but harder. It requires the government to be willing to cut programmes that deliver little benefit. And while the government has taken a sharper eye on the effectiveness of new spending programmes under the social investment approach, too much spending simply carries over, year after year, with little attention paid to whether that spending achieves its objectives.

Interest-free student loans cost the government $600 million dollars per year and mostly benefits students who are either from wealthy families or who are likely to go on to be higher earners themselves. That’s more than what it would cost to cut the 17.5% income tax rate down to 16.5%.

Deciding not to throw $300 million at the film industry over the next four years would allow the government to cut the 30% rate down to 29.5%.

Every billion dollar programme throws away the chance to cut the 17.5% income tax rate to 15.5%.

But, even worse, while the medium-term forecasts are very rosy, with plenty of room for tax cuts, the longer-term projections have health care and superannuation costs requiring substantial tax increases or substantial spending cuts – unless somebody finds the magic formula to reverse the long-term slump in productivity.


New Zealand has an excellent non-tipping equilibrium, but there's been some discussion of encouraging a shift to tipping. A lot of restaurants do kinda have it, mostly (I think) as a way of securing rents from foreign tourists who don't know better.

But most of the discussion around tipping has the base economics of the thing wrong. The main point of tipping is to solve an information asymmetry problem.

Suppose you own a restaurant and simply have no way of telling which of your wait staff are decent with the customers and which are terrible. In that state of the world, you pay them a low hourly wage and encourage the customers to top it up based on the quality of service. Ideally this means that bad wait staff get no tips and exit the industry, average service gets an average overall wage inclusive of tips, and excellent staff are appropriately compensated.

In practice, shame constrains customers against leaving meagre tips for poor service, and tipping makes splitting bills among groups a hassle. And is it really that hard for a restaurant owner to tell whether the wait staff are doing their jobs properly?

Also worth noting is that service staff in roles with tipping have lower minimum wages in the US. Wages plus tips in total have to equal the minimum wage in some states; others just have a really low minimum wage for tipped staff. New Zealand's minimum wage is already very high relative to prevailing wages, and restaurant meals here are not cheap. Combining a high minimum wage with a shift to a tipping norm would have effects on overall demand for restaurant meals.

Finally, and from a more Eric-preferences perspective, it's just really really nice to have the staffing costs bundled into the price of everything and not have to carry a wad of small bills around to pay the same amount for your meal at the end of the night. And I don't particularly like what tipping does to service in the US - because I'm idiosyncratic. On average, people give higher tips where the server is chatty and, if the server is female, touches the customers. I don't like any of that.

Previously: Tipping and inflation incidence

Wednesday, 24 May 2017

Net Fiscal Impact - disaggregated

Keith Ng's 'mythbusting' around net taxpaying annoyed me enough that I started digging around for more of the literature on this stuff. 

Keith argued that it's a myth that 40% of households pay no net taxes. While he's right that the commonly cited measure just focuses on income taxes paid versus cash transfers received, it isn't like correcting for that shows the system is less progressive or relies less heavily on the top deciles' tax contribution. When Aziz et al had a look, they found this:

The figure shows the cost of transfers and services provided to households in each income decile, net of the taxes paid by households in that decile. It includes all taxes, even GST. And when you look at it that way, the bottom four deciles receive substantial transfers, the middle three deciles net out close(ish) to even, and the top three deciles pay rather more in tax than the value of the services they receive. Or, put another way, the bottom six deciles receive more in transfers and services than they pay in tax, the seventh decile is tiny net tax contributor, and the top deciles contribute substantially. 

That should have been common knowledge among folks dealing with data, inequality, and tax. We even included the graph in our report on inequality.

What we didn't include, and I only recently had pointed to me, was the life-cycle and gender effect. Life cycle of course matters: government spends far more on kids than it receives in taxes from kids, and same for the elderly. Much of tax and transfer policy handles life-cycle income smoothing that would otherwise be generally handled by households on their own.

But I hadn't known the size of the gender gap. There had to be one, since men's earnings are higher than women's earnings; women live longer and then collect superannuation over a longer period; and, health care costs are higher for women. Even still, I was surprised. I hadn't known about the gender gap in education costs, for example. 

Regions above the $0 line x-axis are cohorts that, on average, are paying more in tax than they are receiving in transfers and in services (valued at cost). They write:
The data illustrated in the figure suggest that, on average, males start having a positive net fiscal impact—per capita tax revenue exceeds the (allocated) expenditure they receive—in their early 20s. Women, on average, do not pass this “break even” point until their mid-40s. This is due to a combination of lower workforce participation, higher health and education spending, higher income support, and lower direct and indirect taxation.

A possible causal link may lie behind the high value of per capita education expenditure observed for women aged 30–44 and the lagged increase in per capita market income and direct tax for females in the 45–49 year age group. One possible hypothesis is that retraining during child-rearing years that precedes re-entry to the labor market results in an increase in market income and consequently higher direct taxation. The net effect of decreased education expenditure and increased direct taxation increases the net fiscal contribution of women in the 45–49 year old age group.
If we consider only the age patterning, this looks like what households and individuals would be doing for themselves absent the state. You consume less than you earn during your prime earning years in order to provide transfers to your children and to save for your retirement.

There's a similar aspect to the gender patterning where it is more common for males to be the primary earner. In households with a male primary earner and a female who takes a longer period out of work or in part-time work to be able to spend time in the home, you'd see males consuming less than they earn in the market and females consuming more than their market earnings. Were there no government-provided healthcare, taxes would be lower and families would purchase their own health insurance - which would have a similar 'net contribution' breakdown were anyone so daft as to try to produce that figure within a household.

This side of things didn't come in at all for the piece I wrote for the Dom Post on it, as it's largely irrelevant to the argument about household-level net tax contributions. But it's interesting and I hadn't known it before.

Thanks to John Creedy for the pointer.

Also: Fairfax produced this very nice infographic version of the Aziz et al figure. Excellent!