---
title: "29% of Australia's Occupations Are in Shortage. Here's How to Run Selection When You Hire Nearly Every Applicant."
description: "A shortage raises your selection ratio toward 1.0, which collapses the value of ranking candidates but not the value of assessing them. A seven-step redesign, grounded in OECD and Jobs and Skills Australia shortage data and the Taylor-Russell selection tables."
canonical: https://www.assessall.com/blog/29-of-australias-occupations-are-in-shortage-heres-how-to-run-selection-when
published: 2026-11-22
updated: 2026-11-22
author: "AssessAll Editorial"
source: AssessAll
---

# 29% of Australia's Occupations Are in Shortage. Here's How to Run Selection When You Hire Nearly Every Applicant.

A shortage raises your selection ratio toward 1.0, which collapses the value of ranking candidates but not the value of assessing them. A seven-step redesign, grounded in OECD and Jobs and Skills Australia shortage data and the Taylor-Russell selection tables.

A shortage occupation is one where employers cannot fill vacancies at prevailing wages and conditions despite genuine recruiting effort. It is a statement about a labour market, not about any candidate in front of you. A shortage pushes your selection ratio — the share of applicants you hire — toward 1.0. That changes what your assessment is for. It does not remove the need for one.

Most teams respond to a shortage by quietly lowering the bar and hoping. The better response is to change the job the assessment is doing — and to document that you did.

## What a shortage list actually measures

On 14 October 2026 the OECD published [Mapping Skill Shortages in OECD Countries](https://www.oecd.org/en/publications/mapping-skill-shortages-in-oecd-countries_735f9485-en/full-report.html), the third edition of its Skills for Jobs database, covering 38 countries. Its pattern finding is that shortages concentrate in high-skilled work — ICT specialists, science and engineering professionals, health professionals and teachers — while lower-skilled occupations more often show surpluses.

National lists tell the same story with harder edges. Jobs and Skills Australia's [2025 Occupation Shortage List](https://www.jobsandskills.gov.au/news/shortages-ease-gaps-persist-2025-occupation-shortage-list) assessed 29% of occupations as being in shortage, down from 33% in 2024 and 36% in 2023. Nearly half of trade occupations and two in five professional occupations remained short, concentrated in health, education and construction.

Vacancy data points the same way. Eurostat put the [euro area job vacancy rate](https://ec.europa.eu/eurostat/web/products-euro-indicators/w/3-16122025-bp) at 2.1% in Q3 2025 and the EU rate at 2.0%, both down from 2.5% and 2.3% a year earlier.

Read together, the honest summary is that pressure has eased from its peak while a substantial minority of occupations remain hard to fill. But every one of these numbers is a market aggregate. A shortage list tells you the queue outside your door is short. It tells you nothing about whether the person at the front of it can do the work.

## The arithmetic that makes people give up on assessment

Your selection ratio is hires divided by applicants. The classic reference is Taylor and Russell's selection tables (Taylor, H. C., & Russell, J. T., 1939, *Journal of Applied Psychology*, 23(5), 565–578), which show the proportion of hires who turn out satisfactory at a given predictor validity, base rate and selection ratio.

Take a base rate of .50 — half your current people are performing acceptably — and a predictor validity of .50. The tables give:

- **Selection ratio .10** (you hire 1 in 10): .84 of hires satisfactory
- **Selection ratio .30**: .74
- **Selection ratio .50**: .67
- **Selection ratio .95** (you hire nearly everyone): .52

That last row is what makes hiring managers abandon testing in a shortage. If you must hire 95 of every 100 applicants, a good test moves your success rate from .50 to .52, and the effort looks unjustifiable.

The arithmetic is correct. The inference is wrong, for a specific reason: those tables price one function only — *ranking a queue to pick the best of it*. That is the function a shortage destroys. It is not the only function an assessment has.

## Three jobs your assessment still does when you hire nearly everyone

**Establishing minimum competence.** At a selection ratio of .95 you still reject 5%. Identifying which 5% is a real decision with real consequences, and it is a pass/fail judgement against a standard, not a ranking exercise. The distinction matters: a norm-referenced percentile tells you who is best in a thin applicant pool, which in a shortage can mean "best of three people who all fall short."

**Producing a gap profile.** If you are hiring someone who would have failed your old bar, the useful output is not a score but a specific list of what they cannot yet do — an onboarding input. AI-graded scenario assessments suit this well, because they produce reasoning-level detail rather than a single number.

**Creating a defensible record.** Lowering a cut score under hiring pressure gets examined later — sometimes by a regulator, sometimes by your own leadership after a bad cohort. A documented standard, change and rationale are what separate a deliberate adjustment from drift.

## How to redesign selection for a shortage role

### 1. Confirm it is a shortage, not your own funnel

Before accepting a market explanation, rule out the cheaper ones: an uncompetitive wage, a description listing twelve requirements for a role that needs four, internal approval delay, or an application process candidates abandon midway. A shortage you created is fixable this quarter; a national one is not.

### 2. Measure your actual selection ratio

Not the market's — yours, for this role, over the last two quarters: qualified applicants reaching final decision, divided by offers extended. Teams routinely assume .20 while operating at .80.

### 3. Split the requirement into day-one and trainable

Run a short job analysis that sorts each requirement into "must be present on day one" and "can be brought to standard within 90 days." A shortage is a reason to move items into the second bucket deliberately, with a named owner and a re-assessment date. It is not a reason to pretend the items do not exist.

### 4. Re-derive the cut score rather than nudging it

The US [Uniform Guidelines on Employee Selection Procedures](https://www.govinfo.gov/link/cfr/29/1607) state at §1607.5(H) that where cutoff scores are used they should normally be set to be reasonable and consistent with normal expectations of acceptable proficiency within the workforce. That is a standard-setting instruction, and "normal expectations" can legitimately differ for a role you intend to train into.

The same section is useful in another way. §1607.5(G) holds that ranking requires stronger validity and utility evidence than pass/fail use when ranking produces greater adverse impact. In a shortage you are moving away from ranking anyway — which also moves you toward the lower evidentiary burden. The [SIOP Principles](https://www.cambridge.org/core/journals/industrial-and-organizational-psychology/article/principles-for-the-validation-and-use-of-personnel-selection-procedures/730C076329F88919D5C81C6E7D304FE2) (5th ed., 2018) and the [Standards for Educational and Psychological Testing](https://www.testingstandards.net/) both treat cut-score setting as a documented judgement process, not a number inherited from a vendor.

### 5. Re-run the adverse impact check after the change

A changed cut score changes impact ratios, sometimes in the direction you want and sometimes not. Repeat the check on the new rule rather than carrying it over.

### 6. Widen the geography, then verify before pooling scores

Opening a role to a second country is the most common structural fix for a thin funnel, and it raises a measurement question at once: scores from two markets are not comparable just because they came from the same instrument. Establish that before applying one cut score to both.

### 7. Track the below-bar cohort

Record who was hired under the relaxed standard and compare their 6- and 12-month performance and attrition against hires above the old bar. This is the only way the next adjustment is evidence-based rather than another guess.

## The checklist

1. Rule out wage, job description, process delay and drop-off as causes.
2. Compute your real selection ratio for this role.
3. Sort requirements into day-one versus trainable-in-90-days.
4. Re-set the cut score as a documented standard-setting decision.
5. Switch from ranking to pass/fail against that standard.
6. Re-run the adverse impact analysis on the new rule.
7. Verify score comparability before pooling candidates across markets.
8. Assign training owners and re-assessment dates for every deferred requirement.
9. Review the below-bar cohort's outcomes at 6 and 12 months.

Running a pilot on one shortage role is inexpensive to test: AssessAll's pay-as-you-go credits are ₹30 or US$0.50 per assessment, with 100 free credits on individual signup and 250 on corporate signup, which is enough to check whether a gap profile tells you anything your interview did not.

## When this redesign is the wrong move

If your selection ratio is genuinely low — more qualified applicants than seats — none of the above applies, and the Taylor-Russell logic runs the other way: ranking is exactly what you should do, and a more valid predictor is worth paying for. If the role carries a hard regulatory or safety floor, there is no trainable bucket, and the right answer is to leave the seat empty rather than re-derive a cut score. Shortage lists are also lagging indicators, published annually against earlier data.

## Takeaway

A shortage does not make assessment pointless — it changes the assessment's job from ranking a queue to certifying a floor and mapping what to train. The teams that handle shortages well are the ones that make that switch on purpose and write down why.

---

Source: https://www.assessall.com/blog/29-of-australias-occupations-are-in-shortage-heres-how-to-run-selection-when
