Unemployment is down! We Think!

Unemployment numbers are out and they look like good news. Unemployment fell 0.2 per cent to 5.8 per cent. This matters to everyday life in the following four ways.

1. Lower unemployment should mean bigger pay rises, and more options if you lose your job.

2. Lower unemployment increases the chance the Reserve Bank will hike interest rates, which will get you a better return on the money you have stashed in the bank.

3. Lower unemployment tends to force up the Australian dollar, creating cheaper holidays and cheaper imports. (It jumped over US94c this morning for the first time since last November.)

4. Lower unemployment increases the chance of the incumbent government doing well in the polls. “It’s the economy, stupid.”

But…

There’s not total absence of doubt.

If you look at the numbers more closely, there’s a few little things that might make you worried.

For starters, while the latest data, seasonally adjusted, shows the unemployment rate falling sharply, the trend data (calculated over a longer period) actually has unemployment rising.

There are things doubters can always say about the monthly data, that are not especially helpful.

“It’s just one month!”

“The standard error of the unemployment rate estimate for this survey is 0.2 per cent!”

“Seasonal adjustment is not perfect!”

The top two complaints are true and can only be resolved by waiting for more data (which will, itself, be subject to the same issues…).

But the ABS manages seasonal adjustment pretty well. They even account for the fact that Easter was in March last year and in April this year

March is a big month for falling unemployment.

On average, unemployment has fallen in 19 of the last 20 Marches, by an average of 44,000. Why? perhaps it’s only now that businesses are getting over the summer lull.

The ABS takes that effect out systematically. The red lines are the last 20 Marches, almost all of which show falling unemployment. But the average of the blue lines is just about zero:

We can trust this data for now. But it will be very interesting to see if April results in the effect being magnified or reversed.