# Relation $\Delta Unemployment$ and $\Delta GDPgrowth$

I have studied the development of the difference in unemployment and the difference in the growth of real GDP in the Netherlands over the years 1963-2016.

From the picture it seems like a change in unemployment has a positive effect/predicts a change in GDP growth. How is this possible? Has this to do with the fact that unemployment is a lagging variable?

• "From the picture it seems..." Unfortunately all that glitters is not gold. To be frank in my opinion this does not even glitter, the two trajectories seem almost independent. Dec 20 '17 at 12:15
• @denesp I see you're not a fan of Okun's Law.
– EconJohn
Dec 21 '17 at 3:59
• @EconJohn I am not a fan of the law of gravity either if the data does not fit. Dec 21 '17 at 7:07
• @denesp well you may have issues with the variation that exists with gravity around the world. i.e: en.wikipedia.org/wiki/…
– EconJohn
Dec 21 '17 at 17:25

## 2 Answers

Well it seems like you are observing somthing consistent with Okun's Law which reveals empirically negative relationship between Unemployment and GDP.

As Donald Freeman states in the conclusion to his paper titled paper Regional tests of Okun's law

Okun's law is one of the more enduring stylistic facts of U.S. macroeconomics, having withstood numerous tests in the almost four decades since it was first promulgated by Arthur Okun.

• From the Wikipedia page you linked: "The stability and usefulness of the law has been disputed." Dec 21 '17 at 7:10
• You might be interested in this new question by the same user. Dec 21 '17 at 12:47
• @denesp Much like how gravitational pull differs around the world so too Okun's law differs based on region. though the exact estimates vary, a counter cyclical relationship is maintained between Unemployment Rate and GDP growth.
– EconJohn
Dec 21 '17 at 17:24

GDP might be thought of as a product of several factors:

• Labour productivity (gross value added per hour worked)
• Working time (hours worked each year per worker)
• Employment rate (proportion of working-age population in work)
• Age demographics (working-age proportion of population)
• Total population

If you have $E$ working-age people in employment, $U$ unemployed but looking for work and $I$ inactive so neither working nor looking for work, then the employment rate $\frac{E}{E+U+I}$ and unemployment rate $\frac{U}{E+U}$ are related and tend to move in opposite directions

So if other things remain equal (which they do not), an increase in the unemployment rate can be associated with a reduction in the employment rate and a reduction in GDP, and similarly in the opposite direction. This says nothing about causality either way

The point about unemployment being a lagging indicator is that, in some business cycles, employers may sometimes respond to immediate reductions in demand by reducing working time or productivity before they move to dismissing workers, as they do not yet know whether the demand reduction is short-term or medium-term and they want to be able to respond to any quick recovery. This would be an example of other things not remaining equal