Monday, 27 May 2013

Unintended consequences

Once upon a time in French Indochina, there was a rat problem, so the authorities offered a bounty for each rat caught and handed in. Unfortunately, this worsened the problem because the people were now incentivised to create rat farms. And so ideas that may sound good at first can lead to the opposite result intended.

ABC News
27 March 2013
The Australian Council of Trade Unions is pushing for a $30 a week increase to the minimum wage.

"Minimum wage law" is a misnomer. The law cannot force someone to be hired in the first place, so in reality, the true minimum wage is irrevocably $0/hr. Therefore, minimum wage law does not so much ensure more money to society's poorest as it does outlaw all Australian jobs paying between $0.01/hr and $15.96/hr, that people could have otherwise freely entered into. When the state mandates a price floor for labour above the market equilibrium, supply become underutilised and deadweight loss ensues. Unemployment rises, with disproportionate preference towards the most vulnerable (this is why current laws allow lower wages for young workers, trainees, and workers with a disability).

As a parallel, consider some effects of varying the minimum wage throughout the week. The small business owner loses because they cannot afford to pay the weekend penalty rate. The worker who wanted to work on the weekend for less than the penalty rate loses because such employment is deemed illegal. The consumer loses because the business is therefore closed on weekends. If the above three main parties concerned are losing, who then is agitating for such legislation? By reducing the number of non-illegal jobs, the unions win by increasing the bargaining power of their own members.

Monday, 13 May 2013

Sensationalism

Forbes
18 June 2012


A remarkable achievement for a 15-year-old, although no breakthrough in pancreatic cancer management, for most people do not understand biostatistics. While publication in a scientific journal appears to be pending, various online sources have described Andraka’s invention as “over 90% accurate.” Considering the prevalence of pancreatic cancer at 13.66 per 100000, even 99% sensitivity and specificity results in a positive predictive value is 0.0133 in the early detection setting. In other words, even if such a test says you have pancreatic cancer, there is a 99% chance that you actually don’t have it.

UPDATE: Four years after the article was published, I am finishing specialist training in medical oncology and Andraka’s invention has had zero impact on the management of pancreatic cancer.