The UK is not alone in facing an engineering skills shortage that threatens its ability to thrive in a globalised economy and to contribute to solving the problems facing the human race.

Despite growing demand for a skilled workforce, there is an annual shortfall of 59,000 engineers. To make matters worse, nearly a?fifth of the current UK engineering workforce is expected to?retire by?2026, resulting in a .

One reason is the difficulty of attracting diverse talent into engineering. Another problem is retaining it: every year, roughly 35 per cent of new engineering graduates ¨C and most of them are women and ethnic minorities.

The result is that engineering continues to have the smallest proportion of women of all major UK professions; while women make up and just under , they account for only .

³ÉÈËVRÊÓƵ

Increasing the size and diversity of the UK engineering workforce is a complex challenge, but one thing that has been grossly overlooked is the importance of students developing an engineering identity. Research shows that this is an important indicator of persistence in both engineering education and the engineering profession. Who students think they are frames what they think they can achieve and where and with whom they feel they belong.

With colleagues at the University of Bath School of Management, I have studied both students undertaking a traditional engineering degree and apprentices on an engineering degree apprenticeship. Our findings confirm that a qualification does not make an engineer. Some students and apprentices would not call themselves engineers even upon graduation; instead, they would say ¡°I have a STEM degree¡± or ¡°I am a STEM person¡±. They just didn¡¯t feel engineering was a good fit for them. Unsurprisingly, then, they sought jobs in other fields. And they were predominantly female, gay or ethnic minorities.

³ÉÈËVRÊÓƵ

Engineers are not born, they are made. The development of an interest in engineering is shaped by students¡¯ exposure to engineering experiences throughout their education and by the support and encouragement they receive, nurturing an early engineering identity. But a traditional bias, both in the general public and the engineering profession itself, is that males are better at science and technology than females. And even engineering degrees do little to overcome such prejudices.

All the female students I interviewed had already been in the minority during their GCSE and A levels, so they were used to being interrupted and ignored by overconfident male classmates. They would ignore inappropriate comments, saying things like ¡°It's OK: I'm used to it¡±. But they felt that male academics ¨C the majority of faculty ¨C were also overly critical and dismissive of female students, and the cumulative effect of all this was to make women feel insecure.

Instead of focusing mainly on technical content, engineering educators need to acknowledge and address the symbolic aspects of training for the profession. They need to challenge gender stereotypes and to explore what identities are and how they are developed and supported. This could be done by adding workshops, seminars, readings or guest speakers to the curriculum and by actively giving women opportunities to take leading roles in university projects, helping them ¡°try out¡± their engineering identity.

It also means exploring perceptions and interactions. These things are not very tangible (so not very ¡°engineering¡±) but they make a huge difference to the outcomes of engineering education. Faced with a similar problem on its MBA, for example, Harvard Business School found that the predominantly male faculty had unconscious bias when grading the class participation of female students, that the culture did not support junior female faculty, and that women needed to learn to raise their hands in class to make themselves more visible, for instance.

³ÉÈËVRÊÓƵ

Hermia Ibarra, professor of organisational behaviour at London Business School, talks about ¡°second-generation gender bias¡± as the cause of women¡¯s persistent under-representation in leadership roles, and I think the same applies to engineering. It is not conscious bias but it has the same effect. If when I see a woman engineer, I label her a ¡°female engineer¡± and associate with that label ¡°not very good technically¡±, it is likely that I will exclude her from highly technical projects. In my mind, I just want to choose the best person for the job, but I will make the environment feel unwelcoming to her and, over time, she will get the message that she is not wanted and move on.

The challenge is that second-generation bias is invisible in a way that open sexism is not and hence much harder to address. This is why the policies that different organisations in the UK have put in place to attract more women and ethnic minorities to engineering have had such limited success. They are trying to change behaviours, but it is not behaviours themselves that matter, so much as the .

Engineering educators need to do more to help students reassess the meaning they attach to labels such as ¡°female engineer¡± or ¡°black engineer¡±. They must help all students develop an engineering identity. Otherwise, they will continue to mostly train future financiers and management consultants, rather than practising engineers.

Elena Liquete is a researcher at the University of Bath and a senior consultant at CarringtonCrisp.

³ÉÈËVRÊÓƵ