Bryan Caplan  

Does Portugal Show the Signaling Model Is Wrong?

Wal-Mart's Positive Effects... Tear Down These Walls...
Tyler mischievously taunts me on Twitter:
Good thing the Portuguese saved all those resources which Sweden wasted on signaling.
On MR he adds:
In 2009, only 30 percent of Portuguese adults had completed high school or its equivalent, according to figures from the Organization for Economic Cooperation and Development.
Which got me wondering.  What exactly is Portuguese education like?  Some basic facts:

1. Primary and secondary education in Portugal is free.  Attendance is compulsory until 18. (source)

2. Portugal spends a slightly higher percentage of its GDP on education than the U.S.

3. Portugal's PISA scores in 2009 were close to the OECD mean.

Portugal admittedly spends a smaller share of GDP on education than Sweden.  But it's hard to see what on earth Tyler's example is supposed to prove.  Portugal's hardly a case of a country that "invested" way less than normal in education and ended in disaster.  Instead, it looks like a country with roughly normal education policies, roughly normal literacy and numeracy, and moderately disappointing economic performance.

Maybe the reason is that Portugal's policies were bad on other margins.  And maybe Portuguese human capital is subpar for non-educational reasons.  Remember: even in a pure signaling model, low educational attainment remains a symptom - though not a cause - of low-quality human capital (holding education policy constant).

Bottom line: While I don't pose as an expert on the Portuguese economy, there isn't the slightest reason to see it as a walking refutation of the signaling model.

COMMENTS (8 to date)
Bryan Willman writes:

On its face this says nothing about signalling versus human capital, it suggests to me a lack of solid facts.

IF education is free AND compulsory until age 18
AND only 30% of the population has actually finished high school

THEN one of the following is likely true:

a. They have a system which fails like some parts of the US system.

b. The reality is very different from the law, and lots of people don't attend even though school is free and "compulsory".

c. They have some very severe test or rite of completion for a degree that people fail or don't attempt - so they don't have "a degree" but still get normal scores on the PISA and like evaluations.

Tim Worstall writes:

"IF education is free AND compulsory until age 18
AND only 30% of the population has actually finished high school

THEN one of the following is likely true:"

Your list is missing the true explanation, as is Bryan's post.


You have to recall Portugal's history. From the 1920s to the mid-1970s the place was, to all intents and purposes, a Fascist dictatorship. A Catholic, home and hearth for the women one. It was also poor by Western European standards.

Educational spending and standards were very different back then to what they are now: in the inland Algarve where I am it's not unusual to find profoundly illiterate women in the 60s and 70s. It's very unusual to find one in their 20s.

Current educational standards and spending don't tell us all that much about the achievements of the total population for those standards were different in the past.

Scott writes:

I know that anecdote is not the plural of evidence but for what it's worth, one of my very good friends is Portuguese. He speaks fluent Spanish and English and knows a good deal of French. Most Spanish university students don't have anything remotely near his level of English despite years of schooling. He used to be a race-car driver, stagehand and American embassy employee. He's currently a bar owner and a licensed cocktail artist. By all accounts he's a very sociable, knowledgeable and versatile individual. Never finished high school.

To me, it seems pretty evident that he learned everything he needed to know for his professional and personal enrichment on the job, saving the Portuguese taxpayer a lot of money.

Arthur writes:

About the signaling model of education. If the educational system teaches conformity, not just signal it, what would change in your view?

collin writes:

I think Tyler has you on this one. In terms of signaling, I would focus the battle on college not high school. You can throw all evidence against your Jedi Master but Portugal still has a fairly disfunctional economy. Also you preach conformity of lower classes for long term success and where is society going to get conformity without the domaniance of religous institutions. High School both teaches skills and society conformity.

In terms of reviewing history, I am guessing you would have not supported most women finishing High School in 1930's America. In the long run that one seemed to turn out a fairly good decision.

Philo writes:

Resources spent on signaling are not necessarily wasted (*contra* Tyler's apparent assumption): information is valuable, and signaling conveys information.

joe blowe writes:

There is a common phenomena that has not been taken into account in this story.
The same way that when currency controls are imposed in a country and a 'black market' for currency exchange immediately develops, cultural controls such as a very influential Catholic Church or a very strict dictatorship also spurs an 'underground' cultural reaction.
We end up with a country layered along the lines of compliance and non-compliance with the imposed directives and mores. A culture where only the cheater and the corrupt can get ahead. The educated and the uneducated will live side by side.

dha writes:

Do Americans expect Portugal to be fully first world if it had more education or something? It's one of those "European periphery" countries that had coupist dictators and a strong communist party for most of the latter half of the 20th century. Think Spain if the post-Franco reforms had been much less complete. Not that Spain is so great either. There are a lot of institutional reasons it's poor, and I never heard less college education being one of them.

Signalling predicts increased education as a symptom of higher GDPPC, not as a cause, while the correlation doesn't distinguish between the two.

Comments for this entry have been closed
Return to top