Again with the 90-day trial

13/06/2014 § 2 Comments

Several years ago, New Zealand law was changed to allow smaller employers a 90-day trial period for new employees (yes, you wag, ‘smaller’ is measured in number of employees not in centimetres). In the trial period, employers could simply let people go, no harm, no foul. This provision was later extended to all employers.

The Council of Trade Unions did not approve in 2010:

The 90-day trials are part of a “low road” approach to employment. In the 1990s this road led employers to rely on low wages and skills, building a distrustful and ultimately unsustainable workplace environment. It corrodes the trust required for those wishing to take the “high road” of long term, respectful employment relationships which strengthen productivity, skills, work satisfaction, and wages.

According to CTU President Helen Kelly, it still does not approve in 2014:

“The infamous 90 day trial period is a flop. There is no evidence that 90 day trial periods have led to the creation of a single job. In fact it shows that tens of thousands of workers are being dismissed under 90 day trials each year. There’s not a shred of evidence that trial periods have created any additional employment”

The Ministry of Business, Innovation and Employment released a report this week out of its old Department of Labour group. The report concludes that:

Trial periods have provided greater opportunities for workers to be hired.

What the report does not do is assess the impact of the trial period on overall employment. It says:

A key question in this evaluation is whether employers are confident to take on new staff as a result of the trial period provisions. Answering this with certainty would require a detailed statistical counterfactual comparison between firms that did and did not use trial periods, isolating the effect of wider economic and other factors and over a sufficiently long period. Such an analysis is not realistically possible due to data constraints and a lack of an identifiable ‘control group’, and has not been attempted in this report.

The CTU has used this lack to claim that the government has no evidence that the trial has created jobs. And they have no evidence. Because they haven’t looked.

This is a bit of a problem. If one starts from the point of view that labour market flexibility is good to the extent that it provides employees with opportunities and employers with risk-management tools, then THE key question is whether the trial periods led to (a) increased employment or at the very least (b) more employment for traditionally disadvantaged job-seekers. If the rule change doesn’t improve employment outcomes, then it just looks like a shift of power from one group to another with no compensating benefit.

So, really, the key question for the Ministry is, did it work? And the Ministry just shrugged its shoulders and said, ‘Dunno, beats me, it’s too hard.’

We have been here before. DoL/MBIE has already released a report on the 90-day trial period that did not actually answer the central question. And it led 6 months ago to the same wailing and gnashing of teeth — which I discussed at the time.

Let’s put this issue to rest, already. Can I get an econometrician in aisle three for a clean-up? People are spilling out their prejudices and it needs to be sorted out.

Tagged: , ,

§ 2 Responses to Again with the 90-day trial

  • Seamus Hogan says:

    Bill, you seem to be assuming that the econometrician will be able to do the clean-up on aisle 3. But what if the problem MBIE is having is not a shortage of smart econometricians, but an absence of appropriate data. To invoke my inner Tim Harford, the problem was that when the new policy was introduced they did not randomly choose half the country to implement it in keeping the other half as a control.

    • Hi Seamus –

      Two responses:

      1) the law change was rolled out in two steps, so there must be some way to use that to estimate the impacts (which I already did in a simple model), and

      2) we run into these problems all the time, which is what time-series econometrics contends with. Even without a paired comparison, there must be a before-after way to estimate the impacts.

      So, yes, an experimental approach to the policy with good monitoring and measurement a la Harford would have been great. In its absence, we must be able to do better than talk with some employers and recount their stories.

      The LEED dataset must be useful for this. We could examine at the firm level how they were hiring beforehand and then how they were hiring afterward. Or, what about Maori employees? The hypothesis would be that more Maori were ‘given a go’ after the rule change.

      I’m not saying it’s perfect or easy, but can’t we do better?

What’s this?

You are currently reading Again with the 90-day trial at Groping towards Bethlehem.


%d bloggers like this: