What Do CEOs Do With Their Time?

When you start studying questions of how people spend their time, you quickly realize there’s a big problem. Most of what we “know” about life comes from quick response surveys. These are easy to conduct (“How many hours do you work per week?”) but suffer from two flaws: people are clueless and they lie.

The clueless part: Most of us have no idea how many hours per day or per week we devote to different activities. In theory this shouldn’t be a big problem with large samples; the “wisdom of crowds” concept suggests that enough randomness would even out over time, and we’d converge on a number that’s close to a real average. But the problem is that the wisdom of crowds only works in the absence of systematic bias. When it comes to time use, “systematic bias” is another word for lying. We lie about how many hours we work, sleep, spend with our kids, devote to chores, etc. Time feels longer when you’re doing unpleasant stuff (sitting in the dentist chair) vs. pleasant stuff (enjoying dinner with friends). So we recall it differently.

The best way to get around this is to ask people to keep a time diary, showing exactly what they did at any given time. This forces people to face the reality that a day has 24 hours, and a week has 168, and all our activities must fit within these limits.

Unfortunately, there aren’t too many time diary studies out there. They’re laborious (as anyone who’s kept a time log can tell you!) If you’re a researcher with a limited budget, it’s much easier to rely on quick response surveys. But then, we get bad data.

So I’m thrilled that researchers at the Harvard Business School, the London School of Economics and European University are doing a large study of how CEOs spend their time. Much management literature is based on how one CEO thinks he’s spending his time and then blabs about in his ghostwritten autobiography. This isn’t necessarily helpful since 1) his experience may be normal or abnormal, who knows and 2) he’s probably no better at time estimation than the rest of us. He may claim the key to success is working 80 hours a week, but what if he isn’t actually working 80 hours a week? He’s still successful. So what do we make of that?

It’s unclear. But here’s a different question: what if you could track hundreds of CEOs? What if you could look at them for different sized companies, differently governed companies, in different countries? What would you find? Clearly, your average busy CEO of a large company isn’t going to want to keep a time diary, but it turns out that CEO time is seldom a private matter. Most have a dedicated assistant who knows exactly how he’s spending his time. You can ask the assistant, and ask for records, and get a pretty good sense of the picture.

So that’s what the CEO Time Use Project is doing. I interviewed Prof. Raffaella Sadun this morning for an upcoming Fortune.com piece, which I’ll write more about in the future. You can see some results from Italian CEOs here. One interesting finding? Look at table 2A in the appendix for the distribution of number of hours worked per week. It’s a normal distribution around… 40 hours per week. It doesn’t go over 60. The average was around 47. Now, this doesn’t take into account work hours that the assistant was not aware of, such as if the CEO answered emails at night from home, or did some reading on the weekend. But the assistant knows if there’s a phone call booked, a dinner booked, etc. Maybe this is just an Italian thing. Or maybe CEOs don’t actually work 80 hours a week either. We shall see as more data comes out!