One of the advantages of studying the same field for more than 40 years is the opportunity to investigate interesting hypotheses.

But first, a bit of background. When I started collecting vacancies for headteacher posts in state schools in England, way back in the 1980s, an advert in the TES was virtually the only source for jobseekers, so data collection was easy. The exception was the 12 months in the 1980s when News International titles, including the TES, were affected by a strike that prevented publication.

From around 2000 onwards, the TES had both print and on-line vacancies to check for headteacher vacancies. These days the main source of data are the DfE vacancy site, at least for headteacher vacancies. Even so, some schools still advertise in the TES, and a few on regional job boards, especially in the North East region.

My first hunch, not really a hypothesis in the 1980s when I started the work of collecting headteacher vacancies was that faith schools had more difficulty recruiting a new headteacher than other schools, based upon the number of such schools re-advertising. An early analysis supported the hunch, and the challenge facing faith schools is still with us in an increasingly secular society.

My next hypothesis was that most vacancies for headteachers came as a result of retirements. For a number of years, the annual study of senior staff vacancies I conducted for the NAHT validated this hypothesis, at least for headteacher vacancies.

In the days when all schools in the state sector were linked with a local authority, headteachers retiring at the end of the school-year might inform their governing bodies of their decision to retire at the last meeting in the autumn term. The result, a rash of advertisements for headteachers in January, and the bulk of vacancies for headteacher posts were either advertised or re-advertised between January 1^st and 30^th April.  

As a result of the increasing number of schools that are now academies, I have created a new hypothesis. I wonder whether the absence of any real local governing body for each academy might have resulted in headteachers postponing announcement of their retirement to the Trust until a later date in the term starting in January and, as a result, advertisements for headteacher vacancies are now being skewed to the second half of the first quarter of the year, with fewer January advertisements?

Now we are in April, it is possible to consider the data, and compare the data for 2026 with 2006 and 2018 – sadly TeachVac didn’t collect headteacher vacancies in 2016, but at that time just concentrated on classroom teacher vacancies.  

The first thing to note is that there appear to be fewer headteacher vacancies advertised in 2026 compared with either 2006 or 2018.

Year		JANUARY	FEBRUARY	MARCH	TOTAL
2006	TOTAL	509	394	442	1345
2018	TOTAL	358	380	339	1077
2026	FIRST	175	203	289	667
2026	READVERT	13	22	85	120
2026	Total	188	225	374	787

Source Headbase 2006; TeachVac 2018; John Howson 2026

In 2006, the demography of the teaching profession was very different to that of today. Many more headteachers were approaching retirement in 2006, and often decided to retire before reaching the then official retirement age. Nowadays, there is no official retirement age, pension rules have changed significantly, and I suspect the actual number of headteacher retiring each year has decreased.

However, the more interesting piece of data relates to the percentage of vacancies recorded each month. In 2006 and 2018, data collection did not distinguish between first advertisement and re-advertisements during the three months, but counted all but ‘repeat’ vacancies published.

The 2026 data collection exercise I conduct is based upon a collection date, accurate to within three days of a vacancy being published, and a re-advertisement data based upon a new closing date more than four weeks after the original closing date for the vacancy for a headteacher. Re-advertisements with an April closing date have been assigned to March for the purpose of this exercise on the basis of three to four weeks between vacancy being published and closing dates.

Year		JANUARY	FEBRUARY	MARCH	total
2006	TOTAL	38%	29%	33%	100%
2018	TOTAL	33%	35%	31%	100%
2026	FIRST	26%	30%	43%	100%
2026	READVERT	11%	18%	71%
2026	Total	24%	29%	48%	100%

Source Headbase 2006; TeachVac 2018; John Howson 2026

The data does seem to offer some support for my hypothesis, as January only accounted for 26% of vacancies in 2026 (24% if re-advertisements are included) compared with 38% in 2006, and 33% in 2018, when academies were already a feature of the school landscape.

March 2026 accounted for 48% of headteacher vacancies, compared with 33% in 2026, and 31% in 2018. In a future post, I will delve into the issue of re-advertisements that accounted for five per cent of the vacancies in March 2026.

At some point, I will compare academies and non-academy state schools to see whether their advertising patterns for headteacher vacancies in the first three months of the year vary, and will report my findings in a future post.