Why Does More Education Increase Income?

We reviewed the higher education bubble in two previous posts (see here and here).  Today’s post is looking at the fuzzy thinking driving the bubble, as well as the larger issue of which the bubble is merely a symptom.  But first things first – I need to congratulate my Duke Blue Devils for assuming their rightful place atop the college basketball rankings!  And a preemptive apology to my thesis adviser at Northwestern, who may not care much for this post.

At its core, the bubble in higher education is being driven by the belief that a college education teaches skills that are vital to building a successful career, and that higher levels of education are tied to increased worker productivity.  In other words, skills learned in school lead to higher productivity, which is rewarded by employers in the form of higher earnings for the well-schooled.  Here’s the assumed relationship:

In Economics there’s the concept of asymetric information, where one party to a transaction has better information than the other party.  Virtually all job markets have imperfect information.  To cope with this asymetry, buyers look for shortcuts – ways to infer quality – using a process called signaling.  In the case of the job market, employers look at both the level (how much) and quality (where) of education to signal the quality of the applicant.

In my professional career, I’ve seen consistently that higher levels of educational attainment lead to higher levels of income.  This is especially true with workers under 40, who matured professionally as the higher education mania spread in earnest.  We see this relationship in the chart below, based on 2010 Census data:

There’s a clear correlation between levels of education attainment and higher incomes.  The more interesting questions involve causality – does higher education cause higher income, and if so, how, why, and should it?

An Important Distinction:  Correlation vs. Causation

One of the more prevalent learning disabilities in organizations today is confusing correlation with causation, the notion that umbrellas cause rain or diet soda causes weight gain or even that a rock keeps away tigers.  In a professional setting, I’ve seen this confusion lead companies to mistakenly justify expensive marketing campaigns, brand overhauls, even major changes to their strategic positioning.  Let me share just one example.

Several years ago we were asked to optimize the Marketing performance of a large multi-channel business.  The most successful marketing channel was catalog, with more than 100 million copies being sent to millions of customers every year.  Like many direct-response companies, the team tracked if someone received a catalog and then placed an order online or in a store.  Based on that key assumption, more and more catalogs seemed the way to go.  But, we asked, did the catalog really drive the sale, or was it the umbrella that happened to be around when it was most likely to rain?

There happens to be a very easy way to figure this out (though surprisingly few companies do).  We split the customer file into 2 groups.  One group received the normal series of catalogs; the second group received no catalogs.  Here’s what we saw in terms of sales and profit between the two groups (numbers disguised but directional).

Turns out nearly every customer who received the catalogs and ordered would have ordered without one.  The company had mistakenly focused on one variable (received catalog) and seen a correlation with the desired action (bought something).  But when we controlled for the impact of the catalog, we saw it was largely unnecessary.  Could the same thing be happening with higher education and the job market?

The Assumptions Linking Education & Income

We saw that the relationship between education and income is tied through the assumed impacts of greater skill development and then higher productivity.  Turns out, researchers have begun questioning these assumptions.  One study showed that a majority of students gain few to no job-related skills while in college.  A second study (one of many) showed that there was a negligible impact on worker productivity from higher levels of education.  Finally, a Princeton study showed that when you take a student’s inherent skills into account, universities have minimal impact on future earnings.  The first 2 studies challenge the assumption that higher educational attainment causes skills and productivity, respectively.  The last study severs altogether the assumed causal link between educational attainment and higher income.

The chart above makes this point – there are causal factors related to the “level-of-education decision” which are also related to someone’s level of “skill.”  For example, ambition is a causal factor that may lead both to greater consumption of education and to a job-related skill such as hard work.  This is where the signalling function, unable to make such fine distinctions within causal factors, mistakenly attributes a skill to more education.

This isn’t to say that higher educational attainment doesn’t cause higher incomes; it does.  But the assumptions underlying that relationship are wrong.  We’ve assumed that something we can easily measure – educational attainment – is a handy proxy for the thing we really want to measure, productivity.  But shortcuts can sometimes lead you astray.  If we found ways to identify and measure actual causes of productivity, we might, among other things, stop wasting money on unnecessary education.

Why the Higher Education Bubble Isn’t THE Problem

At first glance, what’s driving the bubble is demand from students.  However, the real driver is employers adding, “Degree required, advanced degree preferred,” willy nilly to every job description.  If employers modified or severed the relationship between hiring and college education by better understanding the impact of a degree on what they care about – productivity – we could arrest the dangerous spirals of higher costs and poorer returns within the higher education market.

However, the bubble in higher education is caused by a bigger problem; a misunderstanding of what drives productivity and innovation.  For example, an increasing amount of research (see here and here) shows the outsourcing of manufacturing has a direct impact on our capacity to innovate.  This shift from “sweating” to “thinking” sounds compelling individually, but taken collectively is perhaps the biggest factor driving the oversupply of higher education.  As I mentioned in my recent posts on strategy (here and here), high-level strategy (thinking) is founded on a lot of messy digging (sweating).  If you don’t get your hands dirty (literally and figuratively), you won’t capture the insights that lead to long-term innovation and productivity.

This is an issue increasingly being studied and given attention.  With the direct and opportunity costs we face, the sooner we confront and address these issues, the more confident we can be that our education dollars are being wisely invested, that our kids and grandkids aren’t being saddled with debt they can’t hope to repay, and that the asset that is our workforce – regardless of collar – is properly utilized.