The Reserve Bank governor gave a speech today. There wasn't a lot new in it — the economy's doing OK, wages and inflation are likely to grow slowly and interest rates are going nowhere for ages.
But it was notable for the reminder that orthodox economists sometimes have no idea about how the economy interacts with society, even the smartest of them, of which Philip Lowe is certainly one.
Much of the discussion was around how low productivity was a key factor behind stubbornly low wages growth, even as job creation has remained strong, unemployment dipped and labour shortages are increasing.
It's not just an Australian trend, with other developed countries that have much stronger labour markets also witnessing historically glacial wage rises despite growing labour shortages.
The Reserve Bank would dearly love to see better wage growth, as it would help paper over some of the other cracks emerging in the Australian economy.
"Some pick-up in wages growth would be a welcome development," he said.
"It would help deliver a rate of inflation consistent with the target, it would help with the debt situation and it would add to our sense of shared prosperity."
But how to get that wages growth? The answer, according to the RBA, is mainly though improved productivity.
While no doubt helpful, there are two glaring problems with this solution.
The first is one that Dr Lowe acknowledges.
"Part of the story is likely to be changes in the bargaining power of workers and an increase in the supply of workers as the global economy becomes increasingly integrated," he noted in his speech.
That has meant a large proportion of what productivity gains there have been over recent years have gone to capital (the owners of the machines and information technologies that generated them) not labour (the workers using the new machines and IT).
This is probably partly because workers in Australia have reduced bargaining power since WorkChoices and the WorkChoices-light Fair Work Act that replaced it.
It's also largely because we are in an increasingly globalised market for labour where expensive developed-world workers can be ever more easily replaced by cheaper and often highly skilled labour force in developing nations (or imported through temporary work visas).
So, a focus on productivity without a concurrent focus on the distribution of benefits from that productivity growth will achieve little to increase workers' wages.
Productivity is a very flawed statistic
The second obvious issue is the very measure of productivity itself.
Philip Lowe pointed out that the household services sector has recently lagged all other parts of the economy for productivity growth.
"Output per hour worked in this set of industries is only 4 per cent higher than it was in 2010. In contrast, over this period, output per hour worked is up 13 to 16 per cent in the other industry groups."
But the typical measure of labour productivity — output versus hours worked — is hopelessly inadequate for household services, which accounts for 40 per cent of the workforce.
Why? Because while the second element — hours worked — is easy to measure, the first — output — is almost impossible.
The Australian Bureau of Statistics measures labour productivity by basically comparing real GDP in that sector against hours worked. The "real" means that the value of goods sold is adjusted for inflation to eliminate changes in price and focus only on output.
The problem with this is that the quality of many goods and services changes over time, causing some of the change in price.
The ABS tries to adjust for this, with a reasonable amount of success when it comes to goods.
But the bureau admits many services are a blind-spot for such adjustments.
"For some types of quality change, it is doubtful if any accurate measure of the change can be calculated," it noted in its explanation of how inflation is calculated.
"For example, in the case of services, consider changes in medical operating procedures (eg keyhole surgery) that involve less pain and a speedier recovery, or educational services making a greater use of computers. In these cases generally no quality adjustments are applied."
Service cuts may look more productive
So perhaps the main reason that productivity looks so bad is that improvements to the quality of some services haven't been captured in the data measuring output.
Mandating a higher ratio of early childhood educators to pre-schoolers probably reduces productivity in that sector statistically, even though outcomes for the children have improved by more than the increased investment in labour.
The big danger from this is that worse services may look more productive.
Consider the aged care facility that has 100 patients and 20 staff. If you get rid of 10 staff, but still charge the hundred patients the same price for their care then there's a good chance the statisticians will conclude productivity has doubled.
That could mean a "productive" nursing home is one where patients languish for hours in their own excrement or choke on a meal with no staff around to save them.
Likewise, a productive classroom is one with 40 students to a teacher, a productive pre-school one with 20 screaming children to a frazzled childcare worker, a productive cleaner one who does 20 floors in an office building in a night but only vacuums once a week.
I'm sure you get the picture, and its not a pretty one for the economics profession and its unhealthy obsession with certain statistics.