Afternoon roundup

A closing of some of the browser tabs:

• Eli Dourado on NEPA - America's equivalent to the RMA. Superb seeming primer on how it all works, though I would not be one to notice any holes in it. 
• Road-building is going to get harder, thanks to a court interpretation around water consenting. Minor bits of drainage are going to hit large hurdles. 
• Chris Snowdon in conversation with Marewa Glover on New Zealand's tobacco and vaping regime. 
• Possibly the best approach for "glue ourselves to things" protestors: leave them there.
• Lindsay Mitchell picks up another aspect of the Perry child poverty report: substantial differences across regions in foodbank uptake, sorted by material deprivation. 
• "Rather than address the workforce crisis it inherited, Labour instead resorted to restructuring the health system to put it under greater central government control and make it more vertical. This was a fatal error." 
• NPR's Darian Woods and Wailin Wong on America's blocking Chinese access to higher-tech chips.
• Government may be preferring government-built housing in easing RM constraints. That, as a pattern, will make it harder to get necessary RM reform - if govt can exempt itself from the worst effects of bad RM rules...
• Older piece now, but I want to be able to find it again. Cato's critique of the US Jones Act, which prohibits foreign shippers from transporting goods between US ports. NZ shippers often want this kind of rule here too. It's been doing particular harm in Puerto Rico lately. 
• Sex work legalisation in NZ included measured aimed at protecting temporary migrants from exploitation; those rules may be doing more harm than good.
• Youth justice facilities overwhelmed by numbers.
• Emergency rooms overwhelmed by numbers and want of staff.
• Mental health system overwhelmed by numbers and want of staff.
• One reason the new Rings of Power series may not be as faithful to The Silmarillion's account as it could be: copyright. 
• Wellington: for, by, and of connected insiders. 
• MBIE is running a review of space policy. Wish there were a one-click submission option "Just leave it alone and don't break things." There are just so many submission processes on...
• Yet another area in which NZ could do well by simply recognising standards in use in other places. This time, car safety. 
• It looks like the prescription guidelines for medicines for ADHD have been designed around avoiding Ritalin flowing out into other uses rather than helping patients. And the one doctor who prescribed in accordance to patient needs rather than guidelines is in trouble for it. 

Data integrity and fraud detection

Lindsay Mitchell raises an interesting problem.

In 2011, the Department of Labour matched HLFS employment survey data with benefit data and found:

About 40% of people on work-tested benefits may not be meeting their labour market obligations, as they appear to be either working too much or searching too little.

The backstory Lindsay provides is excellent - it looks like the Ministry tried to bury the paper, and it came out later accidentally. Lindsay only got it with the Ombudsman's intervention. Go read her whole post. And she gives one plausible non-evil reason why the Ministry might have wished to bury it:

But imagine a beneficiary reads or hears about how a survey they are being forced to participate in is being checked against their Work and Income records. For the welfare abuser, that would merely tip them off to lie more consistently to government departments.

Data-matching is being used increasingly but its effectiveness lies in keeping the public in the dark. There's an irony at work. Non-transparency is required to improve integrity of systems.

So ultimately that's where I find the most convincing rationale. But that leaves me with a dilemma.

As a long-time critic of the welfare system, the findings vindicate or illustrate my concerns about the rampant misuse of the system (which hurts genuine beneficiaries and the taxpayers funding it). Do I want to make a song and dance about these findings though, if the information acts to assist those with the worst motivations?

So long as no beneficiary actually was punished for truthfully answering an HLFS or HES survey, the odds of contaminating future survey responses are lower.

The most paranoid end of the distribution would expect that the government has been doing this forever and so always would have lied; the least paranoid end would either expect that the government weren't competent to actually match up records, or that Stats NZ wouldn't be lying about the uses to which their data is put. Without actual cases of "I know a guy who told the truth on the HES survey and *bam* lost his benefit", I wouldn't expect huge effects - but I have low confidence in that expectation.

But I think there's a way around it.

First, link up IRD and HES/HLFS and MSD data from last year through the IDI, along with whatever other administrative data seems useful. Use the IRD and HES/HLFS data to establish true cases of fraud. Use the rest of the data to get the correlates of fraudulent receipt. If the data allows for a reasonable predictive model, great! Save the parameters for next year. If not, abandon.

Then, if the predictive model had been decent, use next year's administrative data to forecast which recipients are at higher risk of fraudulent receipt - and have MSD follow up the higher risk cases. Drop from the sample anybody who was an HES respondent - it'll be a pretty small number anyway. You'll then be pinging those recipients who are similar to last year's fraud cases, but you won't be hitting anybody who was one of the survey respondents. After enough of a lag, bring the prior year HES respondents back in - their back-end data should have changed sufficiently that they won't perfectly predict any more, so they are not being punished for having answered truthfully. They're being audited if their characteristics are still very similar to those of high risk cases.

It won't be perfect - there'll always be some who'll lie on the surveys, just in case. But would there really be many who'd start lying because of this procedure?

Open OIA

Lindsay Mitchell points to some new practice at Treasury: they'll be posting their responses to some Official Information Act requests on their website for everybody, rather than just to the sender, soon after the person making the request gets the information from Treasury.

On the one hand, this is great for open data. Some people already choose to run their OIA requests this way, using FYI, which posts responses to OIAs for those who want the replies to be open.

On the other, OIAs are a bit like gold prospecting. It can take a while to figure out where the paydirt is. The returns to prospecting drop if everybody gets a copy of the treasure map when somebody makes the right request. It shouldn't be an issue if "soon after" means a couple of days. But if it means a couple of hours, it could be.

Lindsay posts one of the OIA results at her blog: GST and excise paid by income decile. Interestingly, the bottom income decile pays more GST than the 2nd and 3rd deciles, and more excise (petrol, alcohol and tobacco) than the 3rd and 4th deciles.

This points to one of the longstanding problems with the bottom decile - nobody much trusts that income in the bottom income decile really reflects anything real. Sole traders with highly variable earnings can be caught in it, for example. That GST paid by those in the bottom decile outstrips GST paid in the next two deciles suggests that there's a fair bit of income smoothing going on. That excise in the bottom and second deciles tops that in the third and fourth points a bit to smoothing but also perhaps to the fairly regressive nature of alcohol and tobacco taxes.

DPB stats

Lindsay Mitchell has a look through the Cabinet Papers for the proposed welfare policy changes.  She points to this bit (see p.2): 

12. A significant number of sole parent beneficiaries give birth to additional children while on a benefit. In 2010, 4,800 newborns (7.5 per cent of all live births) were included in the benefit of an existing Domestic Purposes Benefit client. The on-benefit birth rate has trended up since 1997, from under 35 births per 1,000 women receiving Domestic Purposes Benefit - Sole Parent in 1997 to over 50 per 1,000 women receiving DPB-SP in 2010.

13. As at November 2011, 26,000 women receiving DPB (29 per cent of clients) had included additional newborn children in their benefit at least once since 1993:

• 20 per cent who had a subsequent child included once (multiple births count as one event)

• six per cent who had a subsequent child included twice

• two per cent who had a subsequent child included three times

• one per cent who had a subsequent child included four or more times.

14.  Parents who have additional children on benefit are an at risk group:

• over 90 per cent are single (annual average - see Annex to Paper C for a more detailed breakdown)

• among DPB sole parents, those who have subsequent children on benefit are more likely to:

• have been on the DPB from a young age

• have started on the DPB with a newborn (rather than having exited a relationship with a child)

• have no record of having been employed before, after or during spells on benefit.

About 900 women, while on the DPB, had four or more additional children.

The full set of papers is here. Here's Lindsay's prior report for the NZBR, 2009, on welfare.

I'm not from here, but I'd be surprised if it ever were the intention of those who established the DPB that it be anything but a last rung of support for women and children abandoned by deadbeat husbands from whom payment orders were non-collectable. Nor would I expect strong current voter support for a system that yields these kinds of outcomes.

Lindsay has been pushing this file for ages; part of the impetus for fixing things is due to her efforts. Even getting the stats out seems like a victory.

Welfare reform

New Zealand moves towards welfare reform, with a greater emphasis on moving recipients, especially those on the DPB (our version of AFDC) into work. Those with children aged 14 and up will be expected to make serious efforts to find full time work; those with children over the age of 5 should be finding part-time work.

The change will, hopefully, lead to most income support for single mothers being handled through Working for Families, a wage subsidy scheme similar to the EITC, and indirect subsidies for childcare via both government subsidization of early childhood centres, tax credits for childcare expenses, and additional subsidies for childcare for low income families.

My usual worry on requiring parents on the DPB to seek work when their youngest child hits age X is that it provides reasonable incentive to have another child at that point: if MSD made data available, it would be awfully interesting to run probit on likelihood of an additional birth conditional on hitting that age threshold. Apparently 4,000 children were born last year to women who were already receiving the Domestic Purposes Benefit.

The latest changes have taken some account of this incentive. The birth of an additional child while on benefit provides one-year's respite, followed by a resumption of the status that existed ex ante. So if your youngest child was 5 when the next one came along, the expectation of part-time work would resume after the one-year hiatus. The Greens were worried the government might go farther and require long term contraception as condition of receiving a child-related support package; it looks like National's not going that way. It's not immediately obvious to me why such requirements are so objectionable, but neither is it obvious that the chosen policy alternative won't work. If we're going to worry about fiscal externalities imposed by those who choose to drink too much, or perhaps even ski without helmets, I can't see what logically keeps this off the table.

The changes also incorporate some targeted paternalism. Teen parents and young people on benefits will receive extra payments for completing budgeting and parenting courses; rather than receiving payments directly, their rent and utilities will be paid directly by the government with remaining funds under stricter monitoring through payment cards. Bill Kaye-Blake doesn't like the programme's paternalism:

The ‘managed system of payments’ just sounds awful. It message is, ‘You haven’t learned to take care of yourself, so we’ll just do it for you. Here’s your pocket money’.

I'm no particular fan of paternalism either. But if it's going to be anywhere, targeting it here isn't crazy. Scott Beaulier and Bryan Caplan make a pretty convincing argument that the behavioural anomalies that behavioural economists worry about hold especially strongly among the poor. Note: my post above-linked suffered linkrot; the original Beaulier & Caplan article is here. Beaulier and Caplan argue for reducing the size of the welfare state, but you can also take the article as implying that welfare payments ought to be combined with fairly prescriptive approaches to benefit use combined with directed training to help recipients learn the skills to help keep them off of benefits when time limits hit. Combining the paternalism with training to wean them off the paternalism doesn't sound nuts.

Bill worries too about consequences for existing children whose parents are on the benefit. I'd share some of those worries, but just look at the scope of support already available to working poor families. And payments aren't killed for failure to find a job given an honest attempt; they're killed by proving to a welfare case officer that you're utterly uninterested in trying to find work.

More broadly, the reforms are in keeping with a reasonable social contract for social support payments. Middle class support for welfare disintegrates when it becomes seen as a way of life for those who could work but choose not to. Charles Murray documented that in the US in the 1980s; Bill Clinton's welfare reform followed. Reforming welfare such that it provides temporary support for those falling on hard times, longer-term support for the disabled, and incentives keeping those getting a helping hand from becoming dependent on the system is a pretty good way of ensuring that the social compact over welfare doesn't break down over the longer term. The Greens' complaint that it's unfair  to require DPB mothers to look for work after a year rings pretty hollow for those of us whose kids have been in daycare since they were three months old so we could earn enough to pay the taxes to pay for the system.

Critics also ought to keep in mind that, the longer things are left before reform, the greater the likely support for tougher approaches. American welfare reform brought lifetime maximum welfare eligibility of 5 years for most folks.

Do follow Lindsay Mitchell for what's likely to be the best ongoing coverage of NZ welfare reform.

Bludgers!

I'd previously bemoaned the paucity of data on usage of the various welfare systems in New Zealand: the Domestic Purposes Benefit, the disability benefit, and so on. Long story short: they just don't have records sufficient for any aggregate analysis.

Today's Press has one anecdote, though: one gang-related family that's been on benefit for more than two decades and has received emergency assistance for swimming pool repairs on one of their many properties and for new tyres for their 2007 Chrysler. Farrar comments in horror. But Lindsay Mitchell sees the bigger picture: why can the Press get a full case history on one family when researchers are refused OIA requests on aggregate statistics? It feels a lot more like priming the public to accept some changes to the welfare system than honestly trying to assess the state of the system.

I'd emailed Paula Bennett's office after my last post, trying to get some of the kind of information that Lindsay has been trying to get -- I spend a week in my current policy issues class on poverty and welfare and wanted a better picture of the New Zealand stats. I suggested that if the problem were past records held in paper form, having some summer interns code the data would be pretty useful. Here's the reply I received from Hon. Ms. Bennett:

Dear Mr Crampton

Thank you for your email of 17 August 2009 regarding your recent request to the Ministry of Social Development for information about the lifetime uptake of benefits by beneficiaries.

You advise that you have been told that case records prior to 1996 are only held in paper format. I can advise that the Ministry's SWIFTI system did not exist before 1991. Because of the phased way that SWIFTI built to the functionality that it has today it does not contain full records of people who started their first spell on benefit prior to 1996.

Between the early 1980's and 1991 a much simpler electronic system was used. This system could track a person's interaction with the benefit system at an individual client file level, but could not collate duration information about all beneficiaries, and therefore could not provide 'average duration'.

Prior to the introduction of the first computer system client interactions were recorded in paper files. Most of these paper records will now have been destroyed in line with archiving legislation. For this reason interns completing a 'data entry' exercise would never be able to provide the information that you are seeking.

The Ministry can only really be certain about duration information that it holds for people aged in their early thirties or younger (i.e. those whose first interaction with the benefit system would have been after 1996 when full SWIFTT capture of information began).

I hope this clarifies the situation for you.

When I emailed MSD again asking particularly for data on recipients who entered the system since 1996, I received no reply and didn't have time to follow it up.

We really need better stats on which to base policy decisions. Anecdotes aren't enough.