Agree! 😢

Yes I think it does. We just need somebody to work out how to implement it. As I understand it, the United Learning plan to do this was vetoed by the Treasury.

Thanks Tom :)

Thanks to Gatsby for supporting this and to my brilliant co-author Clare Routledge

Full paper here, including a discussion of the limitations of our approach: repec-cepeo.ucl.ac.uk/cepeow/cepeo...

So what?

1. Starting salary matters for recruitment. This complements the evidence on importance of pay for retention.

2. Recruitment campaigns should emphasise extrinsic rewards, alongside the many meaningful aspects. Great 2015 example here - watch to the end! www.youtube.com/watch?v=vTzz...

Existing research, using traditional self-report methods, consistently finds that intrinsic/altruistic motives dominate the decision to become a teacher: www.sciencedirect.com/science/arti...

So why do we find such contrasting findings? One possibility:

We also find only limited differences in preferences across maths/physics/engineering (MPE) versus other undergrads

The often more severe shortages of teachers in maths and physics is therefore unlikely to be explained by differences in preferences

Drawing on the literature on ‘public service motivation’ we find some differences in preferences for jobs with social impact

In this graph, points further to right indicate more of a preference

But 1) these are modest 2) high PSM grads place no less weight on extrinsic rewards than low PSM grads

It's plausible that those on the margin of teaching (persuadable) are different to the average undergrad in our sample

But we find limited evidence for this among those who report they are considering/planning teaching:

Same graph but showing £ equivalent values:

6 weeks paid leave (grads) -> 13 weeks paid leave (teachers) = +£3.7k salary

Typical 40 hour week -> 52 hours per week (term time teaching) = -£3.2k salary

‘small’ social impact -> ‘significant’ (like teaching) = +£1.2k

TL;DR: Extrinsic rewards matter

Main results in image below

Horizontal axis: change in probability of choosing a job

Eg: Increasing starting salary from £28.5k (grad average UK) to £31.5k (teachers UK) increases probability of choosing a job by 0.08 (8 percentage points)

TL;DR: Pay and hours really matter to undergrads.

We used an online survey experiment with photo-ID-verified respondents

Our sample completed ~20,000 randomised job choice tasks, with values carefully chosen to reflect teaching and non-teaching jobs

Which would you pick?

Reminds me of this UKIP add from 2010. Farage has always known that he functions as the F U button.

Terrifyingly accurate