Interview with David McClelland

One year before his death, aged 80, David McClelland was interviewed in depth for Competency & Emotional Intelligence by Katherine Adams. She spoke to him about his role in the development of competencies and some of the key challenges facing their use by employers. On McClelland see also A summary of McClelland, D. C. (1973). Testing for competence rather than for “intelligence”.

David C. McClelland (1917-1998) spent a lifetime in researching into human motivation. The theory of competency that he developed has been applied to management, small business administration, post-secondary education, mental health, behavioural medicine, economic development and the development of developing countries. He founded McBer & Company and became chair of its board, and was also a Professor of Psychology at Harvard University.

Widely regarded as the father of the competency movement, David McClelland still works as the head of McBer & Company, the consultancy he found in 1963. I talked to him about his ground-breaking work of the late 1960s and his reactions to the way his ideas have been used by other people. I found a man with undimmed enthusiasm for competencies’ potential for good – and some regrets about the way things have turned out.

It was David McClelland’s seminal paper in the American Psychologist1 in 1973 that really started the competency movement in the USA. In that paper, he argued that traditional academic exams and IQ tests were no good as ways of predicting whether someone could, or could not, do a job well. But people tended to believe in them and continued to use them on the basis of the incorrect assumption that the tests measured a kind of intelligence that would promote success in any job. Instead, David McClelland argued we should look for ways to identify other variables – “competencies” – that could predict job success.

But ideas – even revolutionary ones – rarely come from nowhere, and I was interested to know the origins of his thinking. Professor McClelland told me: “I’d been interested in the topic for a while, but I decided it was time to make a fuss about it, because people weren’t paying attention. They were still using academic intelligence tests to do things they had not been designed for.”

“No matter how much psychologists show that some tests work, people don’t use them unless the tests make sense to them”

In fact, he told me, the burgeoning competencies movement of the past 30 years or so has taken a path rather different from the one he originally envisaged. “When I wrote the article,” he said, “I hoped it would be possible to develop alternative types of tests based on an analysis of what it took to do certain jobs. We had a few tentative trial runs, but they never got off the ground.”

The main problem, it seems, lay in the tests’ lack of credibility with McBer’s clients. For instance, McClelland was involved in a project to help the US Information Service recruit diplomats. The government was getting thousands of applications and it was agreed that a screening test would be very desirable. McBer’s researchers identified the outstanding diplomats already in the Service, interviewed them extensively, and compared them with average performers. The characteristics found to identify outstanding performers were then used as the basis for designing or choosing tests.

For example, the researchers found that the best diplomats were very good at divining the feelings of those they spoke to, even when the actual words were giving a conflicting message. So, they chose a test developed by Professor Robert Rosenthal at Harvard that required candidates to listen to “content-filtered” speech, where the words were rendered unintelligible, and to identify the emotions being expressed.

Although McClelland says that this test had moderate success in identifying the better diplomats, it did not appeal to the US government. “No matter how much psychologists say such tests work, people can’t see it,” he says. “The clients thought the test was strange; they preferred to stick to the usual things” – such as test of skill in English-language usage provided by the Education Testing Service, scores on which he says had never been shown to be related to superior diplomatic performance.

What did prove popular, though, was the methodology that McClelland and other McBer researchers had developed to interview outstanding and average performers, and identify their respective characteristics: the so-called “behavioural event interview”. However, compared with pencil-and-paper tests, the technique is relatively expensive, so companies tend to limit its use to work involving top executives and jobs with clear added value, such as sales people. McClelland laments the fact that there still appears to be very little demand for well-researched, reliable and valid competency tests to screen large numbers of lower-level applicants.

In the face of scepticism about testing, behavioural event interviews proved to be the way forward for this client-led movement. Based on an earlier “critical incident” method, the technique involves selecting two samples: people who have been rated as outstanding in their jobs and those who are just average performers.

Interviewees are quizzed in depth about the way they do their work, focusing in particular on occasions when things have turned out well or badly. Transcripts of the interviews are made and each separate behaviour is noted. These “behavioural indicators” are then clustered together into “competencies” that differentiate the two samples. Finally, the competency model is validated using new samples.

For the next decade or so, McClelland and his researchers were busy carrying out interviews and developing competency models – usually, for top managers – for a number of mainly large US organisations. They also designed training programmes to develop the competencies they had identified.

Then, in 1981, the American Management Association commissioned McBer consultant Richard Boyatzis to examine whether a generic model of managerial competency could be derived from the various models that McBer had developed to date. In an influential book2, Boyatzis concluded that there are 19 generic competencies that outstanding managers tend to have – though not all jobs will require all 19, and there are other competencies that may also be required for outstanding performance in any given job.

“There are particular behaviours behind each competency, and you can look them up”

It was Boyatzis who first set down a thorough definition of the notion of “competency” that had underpinned all the work carried out by McBer so far. A competency, he said, was “an underlying characteristic of an individual which is causally related to effective or superior performance in a job”. Competencies can be motives, traits, skills, aspects of one’s self-image or social role, or a body of knowledge that one uses.

In recent years, this definition has been criticised3 for being too broad. I put it to Professor McClelland that by including motives, traits, skills and so on, competency failed to distinguish what was common to all of these things.

He replied: “’Competency’ was a term that was coined to replace the narrower term ‘skill’. In the USA, you can go to a technical high school and learn how to be a filling-station attendant. You practise unscrewing the cap, putting the petrol in, and so on. But, from my observation, that’s relatively unimportant in determining how often a customer stops at the station. More important is whether the attendant smiles or growls at you.”

“So, we developed a competency called ‘customer service orientation’, which covers a broad range of being nice to customers. We call it a ‘competency’ because it’s obviously more than just skills, it covers a variety of types of acts, and each of these gets incorporated into the dictionary definition of a particular competency. There are particular behaviours behind each competency and you can look them up.”

What McClelland is referring to here is McBer’s dictionary of competencies. Developed in the late 1980s and early 1990s by McBer researchers Lyle and Signe Spencer4, the dictionary provides a listing of around 360 behavioural indicators defining 21 competencies that have been found to be most common to nearly 300 different competency models. With each research study that McBer completes, the dictionary is revised and edited to include information that is most predictive of performance.

What McClelland’s argument boils down to, then, is this. Although the definition of competency may be broad, each particular competency is backed up by a list of “behavioural indicators” that define it, clearly and comprehensively.

“A lot of people have jumped on the bandwagon”

Over the past 30 years, McClelland’s work has spawned a huge industry in the USA, the UK and elsewhere in the world, with employers, researchers and consultants all keen to try out his ideas. What are his feelings about this, I wondered?

He told me: “A lot of people have jumped on the bandwagon. The danger is that they may not identify competencies properly.” He fears that many are avoiding the expense of behavioural event interviews by using expert panels instead to “brainstorm” the competencies needed in particular jobs, and then to rate individuals against them.

He says: “Doing it that way is much less expensive, but we’ve shown over and over again that experts only identify around 50% of the competencies you uncover in behavioural event interviews. And expert ratings tend to be poor compared to using the interviews to rate people. It bothers me that companies are doing this without checking the validity of their measures.”

Although some firms are using the McBer dictionary to help their expert panels identify competencies and rate individuals, McClelland fears that this will not be enough to improve the method’s accuracy significantly. Even worse are those organisations whose expert panels simply brainstorm “competencies” without any clear definition of what they mean.

He says: “Firms like these may tell someone they’re rated low in leadership. But when that person asks what this means, all they can say is that ‘whatever people mean by ‘leadership’, you’re low in it’.” Once again, he stresses the importance of having behavioural “anchors” that define each competency in detail.

“The danger with functional analysis is that you leave things out”

Despite the influence of the McBer approach among employers, it is a curious fact that in the UK talk of competencies (or, more accurately, “competences”) is just as likely to refer to National Vocational Qualifications (NVQs) as to McClelland’s idea. This controversial UK government initiative is based on an approach to “competence” very different from McBer’s, and this is reflected in the radical differences between McClelland’s behavioural event interviews and the UK’s favoured method of “functional analysis”.

For where the former is concerned to identify what the people do who are exceptional performers, the latter focuses on the functions involved in particular jobs, and the minimum standards of competence required for each – an approach that has drawn criticism from some quarters.

Although McClelland has not studied NVQs closely, I asked him for his opinion of the general approach. His comments offer an interesting “outsider’s” perspective on this little local difficulty. He said: “It seems like a very traditional way of doing things, not unlike job evaluation, which also looks at the functions people perform and the level at which they perform them. And job evaluation certainly works. But what you don’t find out is the extent to which people who are outstanding fit the model or transcend it in ways that are not covered by the model. The danger is that you leave things out.”

Also of relevance for those concerned about NVQs, McClelland believes that this danger is compounded when the analysis relies on industry experts rather than on detailed interviews with the jobholders themselves. He comments: “When we ask experts to say what competencies are critical, they usually do a kind of functional analysis. But if we follow this up with behavioural event interviews, we find there are gaps [in what they have identified].”

He cites the case of an oil company that wanted a competency model for its business strategists. The experts felt that this was essentially a planning function: deciding what kind of businesses the company should be involved in within particular areas. The experts agreed a list of essential competencies, including “conceptual thinking” and “deductive reasoning”. But the behavioural event interviews revealed that a hitherto unsuspected competency – “influence” – was also very important in distinguishing outstanding from average performers.

McClelland explains: “The outstanding people realised that the job involved more than just writing a good strategic plan. It was also important that top management should understand the plan and be prepared to adopt it. Consequently, the best strategists made sure that executives were involved in decisions at an early stage. The less outstanding people didn’t see this, and it had been overlooked by the experts. But, as soon as we showed them our findings, they could see that it made sense.”

“Our clients have to make up their own minds how they want to organise the company”

Another rival approach that I wanted to ask Professor McClelland about was the “core competence of the organisation”. Though its precise origins are disputed, this idea is becoming influential, especially in the USA, and it has been the subject of several books and articles5.

Originating in the field of organisational strategy, this concept stresses the need to identify the unique factors that can give organisations a competitive advantage. These factors differ from straightforward competencies in that they are seen as an attribute of the organisation itself, rather than the individuals who work within it.

Broadly speaking, core competence is thought to be that combination of corporate characteristics, skills, motivations, knowledge, technologies and systems that distinguishes an organisation from others. It is an intrinsically collective phenomenon, and one that is more than the sum of its parts. In particular, core competence is more than the sum of all the individual competencies of the organisation’s employees, because these competencies are embedded in the organisation’s systems, routines, mechanisms and processes.

Perhaps inevitably, Professor McClelland takes a robustly human resource-centred line on these ideas. He told me: “I don’t see any reason why people shouldn’t use the term ‘competence’ in that way. But I think, myself, that it’s the people in the organisation that are the important thing.”

He described some recent work that McBer has done for IBM, which is now emphasising its “integrated strategy” after years of emphasising the need for each business unit to achieve its own targets. McClelland says: “The new strategy means different parts of the organisation working together so that they can provide solutions to any given problem. But this also means a change for individuals; they have to behave differently towards each other, spend more time with each other. This is the only way the strategy will get implemented.”

The danger is, he feels, that though the strategists’ thinking may get lip service (become “espoused theory”, as Chris Argyris puts it), what actually goes on in the company (“theory in use”) may be very different. He says: “Behavioural event interviews are very generally applicable. You end up with whatever you find – and I think that is important information for the strategists who are saying what the company ought to be doing.”

He goes on: “When we work with a company like IBM, we talk to them about their strategy and look out for behaviours that are relevant to this in the interviews. It might be that what we find out about ‘theory in use’ doesn’t match up with what the strategists want. By identifying the competencies actually being used by a few outstanding people who are already successfully implementing the new strategy, we can let them know more specifically which competencies or behaviours their executives will need more in the future.”

However, he is clear about the limits of the technique. He cautions: “Our clients have to make up their own minds how they want to organise the company. This is not a decision we as HR people can make. What we can say is whether ‘espoused theory’ is leading to behaviour in the best people that will, in turn, lead to that strategy being implemented. But we can’t say whether that strategy is the best. I suppose we might suggest a rethink of the strategy, but I don’t know if this has ever happened in practice.”

“We never deny that a company may have a unique set of competencies”

Returning to the “core competence of the organisation”, however, it does seem that the focus on each firm’s unique sources of competitive advantage has more appeal than McBer’s emphasis on generic competencies. Surely, I persisted, it is unconvincing to suppose that business success for any company lies in getting their employees to do the same things: the things done by “successful” performers everywhere?

Professor McClelland frankly admitted that “we use generic competencies partly to save money for the client, and partly because in many cases we find they are related to success.”

But, he went on: “We never deny that a given company may have a unique set of competencies. We always look for uniques.” Indeed, he sees the concept of competency as a reaction to the traditional search for a single, generalised trait (such as IQ) that would lead to improved performance in all jobs. McBer does not emphasise the generic approach; it is a starting point simply based on statistical analysis.

“We don’t expect any of our generic competencies to be a competency needed for success in every job, by any means,” he says. “Some are relatively rarely associated with success.” Indeed, in a forthcoming article6, he discusses some recent research that examined how often particular competencies distinguished between excellent and average performers in 64 different management jobs.

Although “achievement orientation” was found to distinguish excellence in around two-thirds of cases, in the rest it did not. By way of contrast, “flexibility” – which sounds like an important competency – was associated with better management performance in fewer than one-third of the cases.

“We take it seriously when companies say: ‘We’ll be a new company from now on – we want people to fit the new model’”

McClelland ‘s ideas have been in the public domain long enough to generate a critical literature of their own, with commentators keen to spot the flaws in his arguments. I particularly wanted to raise two contentious issues with Professor McClelland: the “historical” focus of his work, and the issue of equal opportunities.

The first problem identified by critics (7) is that the kind of models produced by McBer are based on factors associated with successful performance in the past.

I put it to Professor McClelland that this makes competencies unsuitable for organisations facing rapid change. He replied: “We worried a lot about this in the early days, but it didn’t turn out to be a serious problem. Companies do often say to us: ‘We’re not interested in this because we’ll be a new company from now on – we want people to fit the new model’. And we take that seriously.”

In such cases, he says, the important thing is to identify those people who already fit the new model, and to study their behaviour very carefully. “Generally,” he goes on, “we’ve found that some of the generic competencies (especially “achievement orientation”) continue to be important for these firms, though others may certainly change. We also now have an extensive database which allows us to benchmark against the leading-edge organisations which have already faced that strategic challenge.”

“American men see power as something you use to correct someone who’s wrong. But women see it as a resource”

The other problem that I wanted to raise was the issue of equal opportunities. McClelland explicitly designed the behavioural event interview technique to identify those factors that accurately predicted job performance and were less biased against women, people from ethnic minorities and other groups.

And yet it seems that there is a whole range of potential equal opportunities problems, some of which are beginning to attract the interest of researchers8. In particular, behavioural event interviews seem open to problems of bias and subjectivity because they are wholly reliant on the judgment of the interviewers and those in the company responsible for choosing the samples to be interviewed.

Professor McClelland told me that he and others at McBer are acutely aware of the problems. “It worries us a lot that interviewers may not be properly trained. We are concerned to ensure that our own interviewers are carefully trained in all aspects of interview technique. Constant vigilance, and sensitivity to equal opportunities issues, should help to mitigate the problems.”

He also shed some interesting light on the debate about whether men and women tend to have different competencies. “We have found that, in US companies, while outstanding women have almost identical competency profiles to outstanding men, there is one area where they differ from men – they are less confrontational and challenging, and this may be a handicap to them.”

“Outstanding American men seem to see power as something you use in order to correct someone who’s wrong, to change them, to show them you see more in this situation than the boss does. Outstanding American women, on the other hand, see power as a resource, something you can use to get people together, to gain commitment.”

Surely, though, I argued, this just underlines the problem with behavioural event interviews. If most of your top executives are men, so that your sample of outstanding performers is also mostly men, would you not be more likely to end up with a competency model that includes the “confrontational” style than the “gaining commitment” style”? And if you use this model to select and train people for top management, do you not run the risk of discriminating against female candidates?

Professor McClelland conceded that this could be a problem. In practice, he still felt that, handled with sensitivity, behavioural event interviews should, at least in theory, allow “alternative” styles to come through into the competency model.

Indeed, he stressed how important it was that this should and does happen: in the USA, there is a clear trend towards less confrontational styles of management. In fact, he said that a recent gender analysis of McBer’s generic competencies found that women demonstrated higher levels and frequencies of key managerial competencies, including both “teamwork” and “directiveness”.

He went on: “We’re studying the glass ceiling at the moment. We’re particularly interested in mentors – people who have succeeded in developing women who do rise to high positions. We want to find out what competencies these mentors have.”

“You can cut costs enormously by reducing interviews from three hours to one hour and not transcribing”

In the decades since McClelland first published his ideas, competency has become one of the most influential concepts in human resource management. At the end of our interview, I asked Professor McClelland to speculate a little about how competency may develop in the years ahead.

His answers show that he is still preoccupied with his failure to gain widespread acceptance for valid, well-researched competency tests that are cheap enough to use for lower-level jobs. He says: “McBer doesn’t have the time to develop them – we have all the business we can handle looking at top jobs. But I still think that this kind of test needs to be developed.”

Another possible solution may be to cut the cost of behavioural event interviews, and here he told me that recent work by Boyatzis may offer some useful pointers. Boyatzis left McBer 10 years ago to pursue the study of competencies in academic life.

His recent book (9) describes his work in developing a competency model that prepares business-school students for managerial success, and argues that carefully designed programmes can indeed enable these competencies to be developed.

Professor McClelland told me: “Boyatzis has cut costs enormously, simply by cutting behavioural event interviews from three hours to one hour, and by not transcribing them into hard copy. Instead, the interviewers are trained to ‘code’ each behaviour with a particular competency straight from the tape. Transcription can be expensive, but coding straight from the tape only takes about an hour and a quarter.”

“When you’re working with top executives, you do want transcripts, because you can’t code for competency levels straight from the tape. But it may be that for lower-ranking jobs his approach would work very well. I’d like to see this developed some more.”

The other innovation that Professor McClelland is pinning his hopes on is computerisation. Today, McBer has upwards of 100 different competency models of senior executives and 7,500 behavioural indicators on around 3,000 executives that are stored on a computer database, allowing the company to carry out a vast range of different kinds of analysis.

Professor McClelland told me: “The database will grow and grow. It already makes competencies more useful to clients because we can tell them how their model compares with those in similar companies across the world. And whereas before we were drowning in paper, with transcripts of up to 90 pages per interview, now we can store the interviews for ever, then retrieve them and rescore them in the light of new information. I see computerisation as perhaps the most important recent development at McBer, and one which has major implications for the future.”

References

1. “Testing for competence rather than ‘intelligence’”, David C McClelland, American Psychologist, vol. 28, 1973, pp.1–14.

2. The competent manager: a model for effective performance, Richard E Boyatzis, John Wiley & Sons, 1982.

3. See, for example: “Competent by any other name”, Charles Woodruffe, Personnel Management, September 1991, pp.30–33.

4. Competence at work: models for superior performance, Lyle M Spencer Jr and Signe M Spencer, John Wiley & Sons, 1993.

5. See, for example: “The core competence of the corporation”, C K Prahalad and Gary Hamel, Harvard Business Review, vol. 90 no.3, May–June 1990, pp.79–93; Competing for the future, Gary Hamel and C K Prahalad, Harvard Business School Press, 1994; Intelligent enterprise: a knowledge- and service-based paradigm for industry, James Brian Quinn, Free Press, Macmillan, 1992; and Built to last: successful habits of visionary companies, James C Collins and Jerry I Porras, Century, Random House, 1995.

6. “Assessing competencies associated with executive success thru behavioral interviews”, David C McClelland, submitted to American Psychologist.

7. See, for example: “Achieving strategic coherence in HRD through competence-based management and organisation development”, Paul A Iles, Personnel Review, vol. 22 no.6, 1993, pp.63–80.

8. See: “Competency: discrimination by the back door?”, Katherine Adams, Competency, vol.3 no.4, Summer 1996, pp.34–39. [NB Competency has since been renamed Competency & Emotional Intelligence.]

9. Innovation in professional education: steps on a journey from teaching to learning, Richard E Boyatzis, S C Cowan and D A Kolb, Jossey-Bass Publishers, 1995.

Note: this interview was originally published in Competency, vol. 4 no.3, Spring 1997, pp.18–23.

Copyright: © Katherine Adams, 1997. Please note this text was published at http://charlie.2-minute-website.com/site/neilrankin.2min/Interview-with-David-McClelland, but that website is currently no more visible. So I (Leonardo Evangelista) decided to publish it here (www.orientamento.it) after trying insuccessfully to contact Katherine Adams. www.orientamento.it is a website managed by Leonardo Evangelista.

Autore Leonardo Evangelista. Leonardo Evangelista si occupa di orientamento dal 1993.