This question sounds like it should have a simple answer, and it is actually a case study in why baseline selection determines the conclusion. Five years back lands on one of the most distorted periods in modern labour market history.
Why the five-year baseline is a problem
The recent past contains, in sequence: a pandemic collapse in employment, an extraordinary rebound with record vacancy levels in many countries, an inflation and interest-rate shock, and then a normalisation. Depending on precisely which month you pick as "five years ago", you can conclude that hiring has collapsed, recovered, or barely changed.
That is not a data problem. It is a baseline selection problem, and it is the single most common way comparisons mislead - deliberately or otherwise.
How to handle it properly
Use multiple reference points. Compare the current figure against:
- One year ago - the recent trend
- 2019 - the last unambiguously normal pre-pandemic year, and the best available "normal"
- A ten-year average - the structural level
- The post-pandemic peak - useful for understanding the discourse, since much commentary implicitly compares against this
If the answer differs across these, say so. "Hiring is below the 2022 peak, roughly in line with 2019, and trending down over the last year" is an honest and genuinely informative answer. A single comparison is not.
Use rates, not counts. Hires as a share of employment, not the raw number - populations and workforces grow, so counts are not comparable across years. See how to make a fair year-on-year comparison.
Use seasonally adjusted series so you are not comparing a January with a September.
Where to get the data
The ONS labour market time series for the UK, and JOLTS for the US, both publish long consistent series you can plot yourself. Ten minutes with a spreadsheet and the actual series is worth more than any amount of commentary, and you will see the distortions immediately once you plot them.
What the analysis typically shows
Two consistent findings when people do this properly for developed economies: the current market usually looks much closer to 2019 than to the exceptional 2021-22 period, and specific segments - entry level, particular sectors, particular regions - are considerably weaker than the aggregate.
Both findings matter. The first means the "collapse" narrative is often measuring the end of an anomaly rather than a crisis. The second means that if you are in an affected segment, the reassuring aggregate is irrelevant to you.
The general principle worth taking away
Whoever chooses the baseline chooses the conclusion. This applies to national statistics, to your company's performance reporting, and to any comparison anyone presents you. The defence is simple and always available: ask why that comparison period, and ask to see the full series rather than two points.
That question catches a remarkable amount of misleading analysis, in business reporting as much as in economics - see the mistakes it belongs to.
The Statistics for Business course at London School of Business UK teaches this kind of critical comparison as core content. Enquire today.