Did you know that keyhole knee surgery does nothing to help people with osteoarthritis?
Subscribe now for unlimited access.
or signup to continue reading
Researchers from America undertook an experiment on a group of people who suffered the condition, and the results were troubling.
The group was knocked out cold with a general anesthetic. Half the group had real knee surgery while the other half had fake knee surgery where their knees were cut open and immediately sewed back together with no actual surgery to the knee itself.
Nobody knew if they had the real surgery or the fake one (and yes, they all consented to the experiment).
The result was remarkable: there was no statistical difference whatsoever between the outcomes of the two groups. Some said it helped. Others said it didn't.
Whether they had the real surgery or the fake surgery made no difference.
Experiments like this produce amazing insights. This surgery used to be performed millions of times a year around the world.
Sadly, economists - particularly macroeconomists - can rarely perform such experiments since major experimental changes to economic policy have major implications for people's lives.
After all, it's hard to justify causing mass unemployment on the basis of economic research. Luckily, we don't need to. Sometimes "natural experiments" - like pandemics, geopolitical disputes and sudden changes in government policies - let us answer a bunch of interesting questions without us having to cause the event itself.
And there have been a bunch of natural experiments lately which, combined with innovative data analysis, have (hopefully) put a bunch of political debates to bed.
Which debates?
Well, if you've ever argued that JobSeeker recipients waste their money, that immigration reduces wages, that immigration is the major driver of housing prices, that Australia is too dependent on China or that working from home is bad for productivity, I've got bad news for you: recent natural experiments suggest you're wrong.
The first natural experiment during COVID-19 was to double the amount of money we pay to unemployed people through the JobSeeker payment.
What did they spend the money on? Alcohol? Gambling? Cigarettes?
Using bank transaction data, we can trace how the people who were already receiving the payment changed their spending behaviour after it doubled.
Turns out, they spent the extra money on groceries, paying rent and paying down debt. There was no evidence that they wasted the money.
There was similarly no evidence that they quit their jobs or worked less.
The second and third political debates are that immigration pushes down wages and pushes up house prices.
Again, COVID-19 delivered a major natural experiment where, for multiple years, we reduced immigration to zero.
What was the outcome? Did wages skyrocket and housing prices collapse? Quite the opposite.
Wages remained just as stagnant, and housing prices went up rather than down, highlighting the dominant effect that changes in domestic demand has on the housing market.
Are these examples of correlation or causation? Countless studies have confirmed that immigration has little to no impact on wages, so the result is unsurprising.
And while increased immigration would certainly add to the demand for housing, our experience during COVID-19 shows that it's one of many factors that drive prices - the biggest being our significant lack of supply.
The next natural experiment to hit Australia was geopolitical. For years, many have claimed that Australia is too dependent on China and that this exposed us to the risk of economic coercion.
Putting aside the bizarre logic that Australia should cut trade with China to stop China from cutting trade with us, the decision by China to impose trade sanctions on Australia was another natural experiment to test this assertion.
The result, again, was to debunk the myth. China imposed trade sanctions covering more than $25 billion of Australia's exports. As far as natural experiments go, this was a big one.
READ MORE:
What was the impact? There was none. Markets did what markets do: they adjusted. The vast majority of businesses found new customers.
While some industries - like wine - were impacted, the data shows that the average effect was nothing more than an inconvenience.
Modelling by our team at Mandala found the same thing. We ran 48 simulations of trade sanctions targeting multiple countries. The argument that concentrated trade with China puts us at their mercy simply doesn't stack up.
The final myth that has been busted is that working from home is bad for productivity. When we sent millions of office workers out of the cities and back into their homes to work, we had all the ingredients for a major natural experiment. And the results are in.
One study of call center workers before and after working from home found that productivity went up 13 per cent thanks to fewer sick days, a quieter work environment and time saved from commuting. They also had higher work satisfaction and a lower attrition rate.
A study looking across all industries found productivity rose 6 per cent.
Another study which used employee monitoring software confirmed the result.
So, there's five long-running political debates which have been thoroughly smashed by natural experiments combined with good quality data.
Whether it changes people's opinions, however, is a different question.
- Adam Triggs is a partner at Mandala. He is a non-resident fellow at the Brookings Institution and the ANU Crawford School.