Cherry picking the RSE evidence

Dr Eric Crampton
Newsroom
8 December, 2020

Decades ago, sociologist Joel Best wrote about how to lie with statistics. The best tricks are those where a statement is word-for-word true but has nothing to do with reality.

Last week, the Productivity Commission released a draft report on companies pushing the boundaries, or “frontier firms.” It’s an important area for study. Low rates of productivity growth mean lower living standards.

But the report’s section on immigration and productivity, and on the Recognised Seasonal Employer (RSE) programme in particular, is a case study in misleading evidence.

The Productivity Commission writes:

“While the scheme has clearly provided benefits for both employers and workers, two recent studies have also shown negative impacts on some RSE workers and their communities (Bailey, 2019; Bedford et al., 2020).”

On first reading, the section appears to suggest that two recent studies found negative effects. That would be surprising. The RSE scheme has been rather extensively studied. Earlier work on it showed it was incredibly beneficial. Could more recent work really show those effects now leaning negative?

The earlier research was conducted by Waikato University professor John Gibson and the World Bank’s David McKenzie who compared families able to send a worker to New Zealand under the RSE scheme with comparable families who could not. What were the effects for both?

Because the research began as the RSE programme was being launched in 2007, the researchers had a before-and-after snapshot. Families were interviewed before workers arrived in New Zealand, and then again at six-monthly intervals over two years. The research design let them draw reliable conclusions about the effects of the RSE programme.

The results were striking. Households that participated saw income increases of over 30%. In poor places, that can matter considerably. It led to real, measurable benefits like building iron roofs for the family home to better withstand cyclones, buying quality household appliances or even opening bank accounts. The families involved said their lives improved markedly.

The benefits of extra income showed up in other ways as well. The older children of RSE workers from Tonga were more likely to attend school. While about 60% of Tongan children aged 15 to 18 in non-RSE households attended schools, having an RSE worker in the household increased school attendance by ten to fourteen percentage points.

In other work, Gibson and McKenzie found RSE workers are far more productive than either Kiwi workers or those visiting New Zealand on a working holiday.

So, it was surprising to read that some recent work might debunk these findings. I was curious what had happened. Had expanding the programme hurt the scheme? Or did something else happen?

The only way to find out was to look up the studies. They weren’t hard to find. But it was difficult to reconcile the actual reports with how the Productivity Commission presented them. I wondered if the draft report’s authors actually read them.

Dr Rochelle Bailey’s 2019 decade-long study, cited by the Commission, does note some negative effects; there are few programmes that have only positive effects. Dr Bailey’s study provides some recommendations for improvements. But her study concludes that “the positive impacts from participation in RSE strongly outweigh the negative ones, which are often unintended consequences of labour migration.”

It mentions the homes built, the community wells dug, and the businesses established due to RSE earnings. It also highlighted how horticulture and viticulture industries expanded.

Dr Bailey’s study provides no basis for cutting back the RSE programme. While it notes some aspects could improve, the context of the programme is strongly beneficial.

But what about the second study? Dr Charlotte Bedford, Dr Richard Bedford and Dr Heather Nunns undertook an extensive stakeholder survey of RSE workers, their families and communities on behalf of Immigration New Zealand.

This study did its best to outline the RSE programme’s strengths and shortcomings. And, like the Bailey report, it concludes that overall, “the findings from the Pacific stream fieldwork are positive for workers and their families who greatly value the opportunity to participate in the RSE scheme.”

The programme is an obvious boost to the workers, their families and their communities. Any negative effects were unintended and often locally specific. As the Bedford study notes, even in places where community members mentioned negative impacts, “no current or ex-RSE worker wanted to terminate RSE employment opportunities.”

Both studies say the RSE programme can improve. But they hardly show evidence of the negative effects motivating the Productivity Commission to recommend scaling it back.

The Productivity Commission frames the RSE scheme as balancing aid with productivity objectives. It sees a beneficial scheme for migrant workers that harms New Zealand. That too is misleading, because the Commission ignores the evidence about the benefits of RSE work for New Zealand.

Bedford et al, the same team surveying the RSE programme’s effects in the Pacific Islands, also surveyed the effects of the RSE scheme for communities here in New Zealand. They found that, rather than displacing local workers or discouraging investment in capital equipment, RSE workers helped expand local businesses. That growth generates “more employment for New Zealand workers” and “large-scale investments in industry infrastructure,” like fruit processing facilities.

The Productivity Commission points to these studies as establishing negative effects of the RSE scheme. It then recommends substantially scaling the scheme back. It is true that each study notes some negative effects. But Joel Best would call that cherry picking. The overall findings are strongly positive.

Framing the studies in this way is far too thin a basis for gutting a scheme that has been evaluated as one of the most effective development assistance programmes out there.

Stay in the loop: Subscribe to updates