Think - AT LONDON BUSINESS SCHOOL

Is there a hidden cost to collaboration?

The consensus on breakthrough innovation tends to be that collaboration is key – but that’s not necessarily so, says Keyvan Vakili

RS1461_LBS-Keyvan--1140-346

In 30 seconds...

  • Organisations are moving to collaborative models because collaboration is seen as wholly positive
  • But individual contributions to projects can be overestimated, leading to a ‘collaboration credit premium’
  • This can incentivise people to get involved in projects out of self-interest, even if their contribution might hurt the project
  • Leaders should look out for the circumstances that can give rise to the collaboration credit premium and guard against doing everything collaboratively 

As they navigate the post-pandemic global reset, companies focused on innovation are moving increasingly towards flat and autonomous structures, inviting cross-functional and cross-organisational teams to collaborate to overcome the critical challenges we face.

Keyvan Vakili, Assistant Professor of Strategy and Entrepreneurship at London Business School, suggests managers might do well to look more critically at the processes involved and the actual value delivered.

When we think about innovation, the consensus tends to be that, to come up with the truly breakthrough ideas, you need to collaborate. You need to pull in expertise, specialised knowledge and diverse ideas from different sources to see the totality of the problem and devise the most innovative, optimal solution. After all, two or three (or 10) minds are usually better than one.

In recent years, collaboration has become something of a go-to approach among forward-thinking organisations. We rely increasingly on autonomous, flat structures where individuals inside and outside the organisation self-organise into groups in order to solve problems.

And that’s a good thing, say many business scholars. In fact, a lot of recent research has discerned a positive link between collaboration and higher-quality outcomes, especially in the creative setting – a win-win-win for collaborators, projects and organisations.

 

Why do we want to collaborate?

But are we right to assume that collaborative work always yields better results in every context? Or could there be a potential cost to it which, because of its very nature, we simply fail to see?

Dr Vakili and his co-authors, Florenta Teodoridis from USC Marshall and Michael Bikard from INSEAD, shine a new light on our understanding of collaboration. Their research study tackles some of our core assumptions about collaboration by focusing on what motivates us to work with other people – and whether the drive to collaborate is just about producing better outcomes for projects, or something else.

“There are a lot of good reasons to feel optimistic about collaboration,” says Dr Vakili. “As we become more and more specialised in different fields, having people co-create ideas and solutions is an efficient way of breaking silos of expertise. When you get different people throwing ideas at each other, you have more of a chance of breaking out of this feeling of combining or recombining existing ideas and seeing something truly fresh and impactful come out of it. There’s also the filtering dynamic that happens with diverse inputs and perspectives: not only do you get the good ideas, but you’re better able to weed out the bad ones.”

Certain costs of collaboration are well understood; including coordination issues, time and language barriers, maintaining commitment at the team level, and the ever-present danger of group think.

 

Individual motivation

Now Dr Vakili and his colleagues have unearthed another, undocumented cost associated with collaboration; one that they believe could have a detrimental impact on project outcomes. It’s all to do with people’s motivation and the credit they get.

“While a lot is known about the costs or frictions that arise in the process of collaboration, less is understood about the incentives people have to collaborate with others and how this might impact or misalign with the prospects that a project might have,” says Dr Vakili.

There’s an assumption that when people choose to work together on something, they do so for the common good and that their collaboration produces better results, which benefits them and the project alike.

While that may be the case, Dr Vakili raises the possibility that individuals may also be motivated to collaborate for more self-centred reasons. Might we not also opt to contribute to a project because of the extra (possibly excessive) kudos or credit we get by doing so?

“We wondered whether collaboration in itself could lead to a situation where each collaborator gets more credit than their actual contribution deserves – and if this ‘collaboration credit premium’ could then incentivise people to get involved in something simply because it benefits them to do so – even if their contribution might not benefit or may even hurt the project.”

The researchers argue that, under certain circumstances, collaborators may indeed receive more credit than their contribution merits, purely because of collaborating and not because of whatever effect their collaboration may have on their output quality.

Imagine two people collaboratively developing a creative solution for a problem and making equal contributions to the final output. Eventually, someone evaluates their output and decides how much each collaborator should be rewarded for the value that they have generated together. Ideally, they would each get a reward proportional to 50% of the value they have created.

But what if they are evaluated and rewarded by different people separately? The challenge here is that the evaluators cannot see the contribution of each person. The researchers hypothesise that in this situation, it is possible that both collaborators receive more than 50% credit for what they have created together, a phenomenon they call “collaboration credit premium”.

In other words, the first person’s bosses may look at their collaborative output and decide that their employee has done most of the work and therefore deserves, say, 70% of the credit for the project.

The same can happen for the other person as well. This way, together, they receive 140% credit for 100% of the work they have done! And they have received this sweet extra credit and reward just because they have decided to collaborate, not because of how their collaboration has contributed to their solution.

Let’s assume the project would have been done more effectively if it had been carried out by each individual alone, rather than collaboratively. The two may still prefer to collaborate because of the collaboration credit premium, even if they know their collaboration would not help the project or may even hurt it, as long as the collaboration credit premium more than compensates for the negative effect of their collaboration on output quality.

"In collaborative processes it’s very hard to evaluate the individual contribution to overall results"

“A good example of how difficult it is to evaluate individual contribution is this study itself,” says Dr Vakili. “We’re three co-authors from three different institutions, and each of us will be evaluated differently – by our own departments and by external groups without a lot of overlap. And that makes it all the more possible and likely that we’ll each end up with more credit for the paper than our actual contribution merits.”

Testing the idea

Testing this hypothesis was not going to be easy, says Dr Vakili, not least because in collaborative processes – particularly creative collaboration – it’s very hard to evaluate the individual contribution to overall results. How do you identify which collaborations are motivated by credit premium when you cannot observe the motivations of individuals? And how do you identify the impact of the collaborations on output quality if you cannot observe them?

This task becomes even more complex when contributors come from different teams or organisations and are therefore assessed differently. To address these challenges, the researchers built on a statistical methodology outlined by Michaël Bikard in a previous study. They ascribed a statistical value or percentage of credit given to contributors in a sample of projects, then simply tallied up the percentages to see if they amounted to a total 100% or not for an average project in the sample.

“The maths is kind of complex, but the idea is simple,” Dr Vakili explains. “The method builds on the assumption that scientists choose a collaboration rate that more or less maximizes the scientific credit that they receive for their papers. By examining how individuals change their collaboration rate across different projects, we can work out the average level of credit that is attributed to a typical collaborator in our sample.

“Using this method, we were able to calculate that, for a typical paper with two collaborators, each individual was on average awarded around 80% of the credit for writing the paper. They weren’t each getting 50%, as one would expect. In other words, every author was being significantly over-recognised for their contribution.”

Next, the researchers used a specific feature of their sample to estimate the effect of collaborations that were formed because of collaboration credit premium on the output quality.

“In the academic field of economics, authors are listed alphabetically on papers. What we knew from prior research is that, astonishingly, authors with last names at the top of the alphabet get a bigger share of the recognition. Those with names at the bottom are routinely relegated to the ‘et al’ division. And in academia, that can have serious repercussions on the likelihood of getting promotions or tenure.”

In other words, economists with names at the top of the alphabet are likely to benefit more from collaboration credit premium, and those whose names are at the bottom are likely to benefit less.

Consequently, the researchers expected that economists with names at the top would collaborate more than their counterparts with names at the bottom. Indeed, they found that moving down the alphabetical ranking by seven letters can lead to 10% fewer collaborations on average.

Built on this finding and using instrumental variable analysis to separate causality from correlation, the researchers then estimated that collaborations that were formed because of collaboration credit premium in their sample had led to lower quality papers on average. Therefore, the authors could confirm that collaboration credit premium can lead to potentially harmful collaborations.

Risks to organisations

The collaboration credit premium effect can pose two significant risks to teams and organisations, says Dr Vakili.

First, because of the difficulty in uncoupling individual input from collective output, it can skew the real value of an employee’s contribution to a project, making it virtually impossible to assess its worth accurately.

Second, the rewards associated with collaboration are potentially so great that people are increasingly motivated to sign up to projects, regardless of how much value they can actually bring. And that can impact outcomes, as well as return on investment.

“It’s hard to assess this risk because no one can see the counterfactual,” says Dr Vakili. “By collaborating, maybe you got the best result possible or a result that was OK, but if you have very little way of knowing the true value of each person’s contribution, you have no way of knowing whether the cost of collaboration was actually worth it – or whether you might have been better off doing something completely different.”

Keyvan Vakili is Associate Professor of Strategy and Entrepreneurship at London Business School