In this page we set out in a little more detail how we measure market tightness for each industry, through the lens of our new modelling tool.
The main idea is that to understand how tight a given industry is — i.e. how hard it is to hire in that industry right now — we need to know two things: Firstly, how many vacancies are firms posting in that industry, and secondly, how hard are workers looking for jobs in that industry?
Our contribution is to develop a method of estimating how hard workers are searching for jobs in each industry. This has to take into account that workers can look for jobs in many industries. For example, unemployed Manufacturing workers might be simultaneously be looking for jobs in Manufacturing firms, while also searching for jobs in Construction.
It also needs to take into account that not only Unemployed workers search for jobs: Employed workers frequently search for jobs and move jobs or industry, and even a large number of Inactive workers — who are supposedly not looking for jobs — find jobs. Workers, especially employed or inactive workers who are not required to look for a job, might also be looking for jobs intensely or not very much at all, so we also need to think about search effort, and how hard workers are searching for jobs.
This allows us to estimate market tightness for each industry at any moment in time as:
Tightness of industry = Vacancies posted in industry ÷ Search Effort directed towards industry
The number of vacancies posted in each industry are directly measured by the ONS, so our contribution is to find a way of measuring the Search Effort directed towards each industry. We do this using the Labour Force Survey, by measuring the number of people who find jobs in each industry. Essentially, with a little bit of work we can work out the search effort people must have directed towards each industry in order to explain the number of people who found jobs in each industry.
The idea is simple at its core. Take the group of people “unemployed workers whose last job was in Manufacturing” in a particular quarter. If many of those workers found jobs in the Accommodation & Food sector in the next quarter then they must have been putting some search effort into looking for jobs in that industry. The more workers find a job there, the higher the search effort put in.
This is complicated slightly by the amount of vacancies being posted: maybe these workers weren’t looking particularly hard for jobs in Accommodation & Food, its just that firms in that industry were posting a large number of vacancies, making it very easy to find a job there! This is where the model comes in: it uses the concept of a “matching function” to balance the role of search effort and vacancies in explaining why workers found jobs in each sector. This allows us to back out the search effort and direction of each group of workers, using data on their observed job finding flows and the number of vacancies in each industry.
The working paper with all the details is just being finalised, and will be posted on our Findings page in the coming days. In the meantime, slides explaining the methodology and results are already available here. So if you would like more information on the details, please stay tuned, or feel free to contact us.