Deciding to proceed

At the end of this level you make one decision: this posting, or a different one.

There are three honest outcomes and only one of them is failure.

Proceed. You meet or can bridge the load-bearing requirements. The gaps sit in the administrative requirements or in areas you can close during the course. This is the common case and it is the one the rest of the course is built for.

Proceed and close a gap in parallel. One load-bearing requirement is genuinely missing and is closable inside fourteen weeks: a named software package, a specific reporting standard, a certification with a short lead time. You continue the course and you spend an hour a week on the gap, and by Level 4 the gap is a sentence about what you did rather than an absence.

Change the posting. Three or more load-bearing requirements are missing with no bridge, and closing them is a matter of years rather than weeks. The correct action is to find a posting one step closer to where you currently stand and run the course against that. This is not a failure of the exercise. It is the exercise working, in the two weeks it costs rather than in the six months of unanswered applications it would otherwise have cost.

The failure mode is the fourth outcome, which is to proceed while privately knowing the third one applies. It produces a course spent building evidence for claims you cannot support, and it terminates in Level 5 with a panel finding the seam in about four minutes.

The facilitator principle behind this course states the test cleanly, and it is worth adopting as your own: for every claim, ask what evidence makes this statement true. If the evidence is weak, improve the capability. If the evidence exists, communicate it better. If the role is beyond current readiness, work out what has to happen next.

That principle sorts every problem in the rest of the course into one of two kinds, and the two have opposite treatments.

A communication gap means the capability is there and the evidence for it is not visible to the employer. Labour economists have a better name for this and a much larger literature behind it: an information friction, in the tradition that began when Spence (1973) modelled education as a signal precisely because productivity cannot be observed at hiring.

The cleanest demonstration that the gap is informational rather than one of capability comes from New York City. Researchers provided credible skill signals, in the form of recommendation letters based on supervisor feedback, to a random subset of 43,409 participants in the city's summer jobs programme. Letters raised employment the following year by 3 percentage points, a 4.5 per cent increase, and cumulative earnings over four years by $1,349, a 4.9 per cent increase. There was no evidence of increased job search or increased confidence. The signals worked by helping employers identify successful matches (Heller & Kessler, 2021).

The authors state the point in their own words: information frictions constrain employment "even when an applicant has the appropriate skills to succeed in a job".

Read the null alongside the effect. The letters did not make anyone try harder or feel better. They changed what an employer could see, and that alone moved employment and earnings four years out.

Levels 2, 3 and 4 are entirely about information frictions. Most job-search advice is written as though every gap is one. Most of the anxiety people bring to job searching comes from suspecting that some of theirs are not, and having nowhere to put that. Putting it in the third column, in writing, in week two, is cheaper than discovering it in a panel.