A couple of weeks ago, I’d talked about when to sell your creative ideas, based on a paper by Dr. Hong Luo at Harvard Business School. This week, I’m looking at more of her research, this time on how break times help college students innovate.
The paper, Slack Time And Innovation, is available for free, and is eminently easy to read, and I encourage you to check out the data.
Why Slack Time?
The paper opens with a Paul Graham quote:
“It’s no coincidence that Microsoft and Facebook both got started in January. At Harvard that is (or was) Reading Period, when students have no classes to attend because they’re supposed to be studying for finals.”
It sounds like anecdotal pattern matching, but that’s basically what science is supposed to do - investigate empirical observations.
It rings true to me - I started my last project when I had a few days off of work, and I got pretty far with it in that short week. I also took a break for three months earlier this year, and that’s when I started my current novel-in-progress, San Francisco Serial.
Apparently traditional models of innovation disagree. These models apparently consider the marginal benefit of each project to be the marginal cost of producing it, and hence, working on a project in ten discrete time intervals is the same as working on it in one large block of time. So you’ll come up with great ideas on your break, but only because you have more time. And by this model, if you are innovating at work AND on your breaks (wait, what’s a break then?), you’ll pick less awesome ideas on your break, because you value your work time a lot, and make sure to go for the best value for time.
But then, there’s companies like Google and 3M, where there are divisions entirely dedicated to innovation, but those employees are also offered generous leave, sabbaticals, and other breaks, because they apparently boost innovation. If work innovation was the same as break innovation, these would probably not be a thing. I feel that’s a shaky thought, as breaks are also for recharging your batteries and exposing yourself to new ideas, but okay, let’s go with it.
Anyway. Since we know that interruptions ruin the quality and focus of your work, the hypothesis is that you should be able to innovate better when you have extended periods of time. Like, during breaks, since, obviously, work is full of interruptions, with standup meetings and design reviews, and Karen’s going away party.
So just like in the selling ideas paper, they come up with a theoretical model, then try to see if it fits some real world data, and then have a bunch of takeaways now that the model has been proven true.
In this paper, their model assumes that during the work week, you value ideas less depending on how complex they are, because you don’t have big blocks of time for them, and make a few predictions based on that model. Then they look at data from Kickstarter, for 165,410 projects on Kickstarter between April 2009 and April 2015, and see if college break times (summer break, winter break, spring break, Thanksgiving, snow days, reading week) correlate with higher number of projects.
TL;DR
People come up with way more ideas during breaks.
During breaks, people choose to work on complex ideas with longer time horizons, and more collaborators.
The ideas that people come up with on breaks are more financially viable than the ones they come up with on non-breaks.
The Model
Let’s say there’s two kinds of weeks - a break week, and a work week (college work, not gainful employment work).
When you have an idea, it has a value you attribute to it when you think of it, let’s call it Q, and an apparent complexity, let’s just refer to it as complexity (k in the equations you’ll see).
When it’s a break week, you value the idea at Q.
But if it’s a work week, you aren’t able to focus as hard, and your friends are busy with their own schoolwork, so you are less confident about completing more complex ideas, so instead of valuing those projects at Q, you value it a little less. So something like Q/e^(𝜷*k), where the more 𝜷 is, the more you value uninterrupted blocks of time. So if 𝜷 is 0, you are a Masai warrior who can work at any time, sleep at any time. And if 𝜷 is 1, you’ve got superADHD with extreme hyperfocus.
If you’re not particularly math savvy, it’s okay, you can simplify it to be Q/(𝜷*k) in your head. No weird exponents. All we want is to say the more complex an idea is, and the more you value big blocks of time, the less you value the idea when you’re busy. Cool? Cool.
And then every idea takes a certain cost (time, money, number of texts on your shared Whatsapp group) to complete.
So,
Payoff during break = Q - Cost_During_break.
Payoff during school = Q/e^(𝜷*k) - Cost_During_School.
Let’s see what effects this has.
In case you’re a Masai Warrior (𝜷 = 0), the difference between break payoff and school payoff just boils down to how much you value your time during break vs school. Since time during school is way more expensive (because you get maybe an hour a day), you will pick the projects with the most payoffs and do them.
But if you are a regular person, (𝜷 > 0), your Q value itself is dampened during school. If you have a test coming up next week, you’re probably going to think your idea for the next Facebook isn’t that important. So you’re probably not going to pick complex ideas, unless you know it’s going to pay off big, because otherwise, the payoff just doesn’t seem worth it.
Based on this, the model predicts three things:
Prediction 1: Probability of developing an idea is higher during breaks. If you’re a Masai warrior, you’d not mind complex ideas during the school year, but if you’re a regular person, you probably filter them out.
Prediction 2: If you’re not a Masai warrior, on average, you work on more complex ideas during breaks. This is because during the school year, you’re filtering out the too-complex ideas, so on average, the complexity of the idea you work on is lowered.
Prediction 3: If you’re not a Masai warrior, you work on more valuable ideas during breaks. This is because you’re filtering out ideas you consider too complex during the school year, so on average, you don’t work on things that are as valuable, whereas on breaks, you aren’t affected by complexity. This isn’t just because you have less time available in total. This effect is worse the more you value large blocks of time (Super ADHD). But if you do happen to be a Masai warrior, you work on more valuable ideas during school than during breaks.
So if the data can fulfill these three predictions, it is proof that we are all not Masai warriors, and that we actually value the large swathes of times offered to us by breaks.
The Data
The researchers pulled data from 165,410 Kickstarter projects that were listed between April 2009 and April 2015. These projects have team bios and target amounts. 44% of the projects succeeded and got 89% of the total money. Success was defined as the project meeting its target financial goal.
Also, in 2012, Kickstarter imposed new rules on the projects in their Design and Tech categories, after some people raised a lot of money and then failed to deliver. They were supposed to share the track records of the team members, their manufacturing plans, and also have a prototype. This information was used to indicate complexity.
They also pulled school break data, which was dates for summer, spring, winter, and Thanksgiving breaks for 200 universities in the US, as well as Reading week, and snow days.
It’s interesting how they correlated this data. What they did was pick out the location from the Kickstarter project, and if there was a school within 5 miles of that location, they connected the location to the school. So a project in Merced would be correlated with the break times of, say, UC Merced.
The expected time for fundraising was considered an additional measure of complexity (k), and the amount of money the goal specified was considered a measure of value (Q).
The Results
So let’s see how they prove the predictions of the model.
Prediction 1: Mo’ break, mo’ kickstarter.
They found this is indeed true. Whenever a location had a break, there were more projects listed on Kickstarter on those weeks, and more money flows to that city.
But then, are the increased projects caused by breaks? They look at 4 ways to prove this is so:
Absence of pre- and post- trends: They marked a baseline that was five weeks before a break. There was no significant drop before the break, or after. What’s more, there wasn’t a spike before or after the break either, in comparison to this baseline. So it isn’t that the students have these ideas anyway, and just choose to work on them during the break. Further, if they were posting during the breaks because that’s when people feel generous, there would be a drop prior to the break. But that’s not the case. It isn’t a shift that is observed; it seems like the break itself is a cause for the idea.
There’s an increase in the Kickstarters of the type that the schools on break are: If it’s a top tier engineering school that was on break in a location, there was an increase in the number of tech projects. If it was an art school, the increase was in art projects. And so on.
Funding does not increase during breaks: They check to see if projects posted 8-14 days before the break experience any spike in funding during the break. It doesn’t.
Snow Days: These are unexpected breaks. Yet, the effect is the same as it is on the expected break days. There’s more projects posted on snow days. On years with more snow days in a location, there’s more projects posted as compared to years with less snow days. Another reason to move to Boston/Cambridge, and another reason to fight climate change.
Prediction 2: Mo’ Break, Mo’ Complex
The data needs to prove that the projects posted on breaks are more complex than those that were posted during the school year, in order to show that we aren’t all Masai warriors, and we actually benefit from breaks.
One measure of complexity is team size, simply because of the overhead costs involved in coordinating between many people. They found that on average, teams posting Kickstarters on break had 0.5 people more than teams posting during the school year.
Another measure of complexity is expected reward time, i.e. the time horizon for fundraising. Projects with longer timelines were posted more during breaks. There weren’t any pre- and post-trends with timelines and breaks.
Another measure is to just see how hards it is to get the project to post on Kickstarter. They looked at the post-2012 data in Design and Tech categories, where the requirements became more stringent (requiring prototypes and plans and team backgrounds) and hence required more effort. These projects went up disproportionately during breaks.
Prediction 3: Mo’ Break, Mo’ Money.
The data needs to prove that the average value in break weeks was higher than in work weeks.
Here, they consider the target amount of the Kickstarter to be a measure of value. Further, they only look at successful Kickstarters that achieved their funding goals, because if people paid that much, it probably wasn’t overpriced.
And they found that for successful projects, the amount they obtained in breaks were much higher than during the school year.
But what if this was because a disproportionate number of projects failed over the breaks, and there was a heavy survivorship bias? To discount this, they looked at failure rates over time, and break times didn’t have a disproportionate number of failures over the school year.
Further, to make sure that the money indeed depicts the value, they looked at the post-2012 Kickstarters in Design and Tech, where the efforts had gone up to post a Kickstarter. The target amounts increased, disproportionately compared to other categories, so though it’s an indirect proxy for value, it can be considered to correlate with value and effort.
Conclusions: Now what?
This model works under the assumptions that work times are full of interruptions, and break times are full of uninterrupted time, when all your friends are free. How true is that IRL?
Grownup Vacations
My biggest gripe about every vacation I’ve taken since 2009 is that I need a vacation to recover from my vacation. My breaks, especially if they involve travel, or family, don’t give me uninterrupted swathes of time. I don’t think I’m alone, and I suppose this gets much worse for people with young children, as during vacations, they need to supervise their child fulltime, while during the school year, at least the kids are at school some of the time.
So if you have a company full of young parents, they would certainly benefit from longer vacations, but it might not lead to more innovative ideas prototyped. That said, my friends with children often say they are full of ideas, owing to the mindless play they indulge in with their children, but are strapped for time to implement them. They do really well as managers, professors, and other leaders.
One messed up story I heard was of a guy going on paternity leave, but he barely changed a diaper as his wife was a SAHM, and had a lot of help from their parents. Instead, he used that time to found a new startup, and test the waters to leave his job.
There evidence of how male professors use parent-friendly policies to further their own career, while their female colleagues use the extra time to recover from childbirth and tend to their infants.
So maybe the right approach is to have a weeklong working retreat? I know some companies offer that for upper management, and some startups do that as well. We all know Bill Gates takes off to some cabin in the woods for a week every year to come up with new ideas. It could just work, having a combination of relaxation, with all possible collaborators already available, and the ability for each employee to structure time without interruptions.
ADHD and Side Projects
As someone with executive functioning issues, interruptions throw me off much more than they do to the average person. So my 𝜷 is rather high. I would love, love, love more uninterrupted stretches of time, because my head is always buzzing with ideas for side projects. This study gives me a little more confidence to plan for week long breaks to work on my side projects, and focus on my work the rest of the time. I realize it’s untenable to try to do both. If I can put in more deep work to produce better results and better value, I should probably do that instead of a lot of shallow work.
That said, I notice I don’t pick less complex projects to do on working days. I do end up working on things that are the most important to me. If I were to take a break, I’d probably work on the same side projects as I’m doing now. But, it’s also that I’ve completely given up on projects that would require more collaboration. I’ve given up on improv and sketch comedy, and screenwriting because they are untenable on my schedule. If I was making short videos and podcasts with friends, it would probably be way more profitable than just writing fiction and nonfiction
Why doesn’t this work for NaNoWriMo?
No Plot? No Problem by Chris Baty, the founder of National Novel Writing Month (where you try to write a novel in 30 days) advocates that you don’t quit your job to write your novel. Instead, just make it your main side project along with your regular life involving a job and kids. The book suggests you have a clear goal of 50,000 words, plot in advance, and have a supportive community to help you achieve your goal.
Early in the book, Mr. Baty talks about how he actually took time off to write a novel, and found that he was unable to stay on track, and instead procrastinated by doing needless chores, and his plot never progressed. According to this model, having a break should be the perfect way to write a novel, shouldn’t it?
But maybe the assumptions don’t hold?
With pre-plotting, you reduce the complexity of the idea, so maybe that helps. But additionally, you don’t value your idea less on working days. In contrast to this model, the culture around NaNoWriMo makes sure your Q is higher for working days, because you are making it a clear priority, and meeting your daily goal is all you have to do.
Additionally, internal interruptions (i.e. distractions) are plenty when you have a long period of time and an open-ended task, so there’s probably an upper limit on how much uninterrupted time is good enough to get things done in. Plus, there’s daily, weekly, and monthly goals, which quantify creative work. The quality doesn’t matter, so if you can steal moments here and there to make your goals, it’s still as good as having an uninterrupted time block to do the same.
Maybe we have got to come up with a model for NaNoWriMo, make predictions, and see how it generalizes with the large number of people who take part in it.
Conclusions - For Real This Time
This Onion article hit me right in the feels. My husband and I work fulltime, and then come home to work on our side projects. We work hard to keep up with our passions, and often, I wonder if it’s worth it. We want to have great ideas, and implement them, and feel the joy of accomplishment in something we are passionate about, and we are always desperate for long stretches of uninterrupted time.
That’s a near impossibility, because, well, life, but we manage. I wonder how applicable this model is to us.
It’s possible that our side projects are a second job to us, which is why after a point, we don’t take risks anymore and just fight to stay on track to complete the projects. It’s also why whenever we get a few days off, we try something new, and now we have one more additional project to procrastinate on.
Tina Fey said in her memoir that being Head Writer at SNL, and then the creator of 30 Rock, was not about encouraging creativity in her team, but about policing it to be consistent with her vision. Maybe, for the small guy with big dreams, that’s good advice. At our level, ideas are cheap, and implementation is everything. As creative individuals, we’re each a one-person SNL writing room. Maybe just picking an idea, and then sticking with it through thick and thin and seeing it to completion is the way to go?