Time capping can be defined as fixing the number of hours for a certain task, research project or decision and keeping our research within those bounds. Most tasks can be completed at different levels of depth and research itself is never-ending - a single topic could often be researched in an hour or could equally have an entire PhD made out of it. The same can apply for website design, outreach, polishing or many other tasks that an organization engages in. Given tasks that are not time capped, people will generally spend more time on doing what they find fun or what they get absorbed in instead of what is best to put hours into in the long run. By setting a time cap on a task we are pre-determining how important that task is relative to other counterfactual tasks. This approach often results in more getting done at some cost of depth, as often 90% of the value of tasks is captured by the first 10% of the effort .
There are a few specific kinds of tasks that benefit especially from time capping. Research and decision making are tasks that frequently result in endlessly flexible deadlines without a clear ‘good enough’ point.
Research is one of the foremost areas that can benefit from time capping. At Charity Entrepreneurship we allocate a specific time frame for every research project. This has allowed our team to cover a lot of ground in a predictable way, and to cover the ground we need for making a specific decision within a clear time period (e.g. conducting research for a year and then recommending a list of top charities that could be started).
Many research teams could think of situations where twice as much research, but with half the time put into each piece, would likely result in more good for the world. For a more concrete example: we had a very specific time budget for our animal reports. For each of the reports we could have gone much deeper (we ended up spending 1-5 hours per report and published a 1-page summary for 15 animals). But it would have been at the cost of breadth: for example, we could have spent 2-10 hours per report, but only covered 7 animals. Or, in theory, covered 30 animals at half the depth. The key questions to consider were what the purpose of these reports was and how big a role did they play in our endline goal (starting effective charities). What level of depth would give us enough information so that we could start to compare and consider which animals would be a priority? We decided that 1-5 hours would give us enough information for the soft prioritization of which animals to focus on.
I think highly intellectual cultures (such as the EA movement) tend to undervalue this sort of time-capped research approach, and I often see comments on research, even very deep research, that more or less translate to “put more time into this research”. Of course, sometimes it is true that more time should be put into a specific branch of research, but I very rarely see it the other way around - comments that roughly translate to “you should have put less time into this research”. It's become a cliche in academic research to say “more research is needed,” but in some areas this is really not the case, and particularly so once counterfactuals are taken into account. Time capping, I think, is the one step that can be used to improve the situation. If someone wants to make the case that we should have covered 7 animals, but at double the depth, I am very interested to hear the considerations in favour of making this tradeoff, but if they say that more hours put into the research would make it better without a thought towards counterfactuals then it’s harder to engage with.
Decisions are infinitely complex and, much like research, it would be easy for a person to spend anywhere between 1 minute and multiple years to consider complex decisions. Some decisions are sufficiently complex that the perfect answer can never be found; just better and worse guesses. Given the uncertainty and complexity of the world, but also the importance of making many decisions (often hundreds a day), time capping the more lengthy decisions seems like an optimal solution. This stops analysis paralysis and constant indecisiveness. It also gives important decisions a fixed deadline and timeline for when they need to be made.
Using an example within CE: at the end of the year, we ultimately have to choose what are the best charities for us to recommend. This decision will almost by definition always be incomplete and subject to revision. But like with the research above, the important question is what are the counterfactuals? If CE, for example, did 5 years of research and decision making in the animal space, would it result in higher impact charities recommended at the end? Almost definitely. But we also have to consider the impact of a strong charity, or two, that could have been started 4 years earlier, as well as the cost of research. CE could cover and recommend charities of mental health, poverty, far future, and meta science. If 2 charities came from each of those areas during those 4 years that would be 8 charities not started for the end benefit of a better animal recommendation.
Some decisions are so complex that it's easier to have impact in multiple fields (or double the impact in a single ‘best guess’ field). Our team ended up thinking that it would be quicker and higher impact to run a year of CE on both animals and poverty issues than it would be to research and decide which one is of higher impact.
This article has also been published at CharityEntrepreneurship.com
If you want to be the first to know when we will start accepting applications to our incubation program, subscribe to Charity Entrepreneurship's mailing list.
I think that time capping can sometimes be the right decision but I’m also afraid of people overapplying it in research. I remember how I was writing Is the percentage of vegetarians and vegans in the U.S. increasing? I did a couple of days of research, concluded that veg*ism is trending upwards, and planned to share my findings in a short post. I was hoping to finish it in a week. But when writing it up, I realized that some of the evidence was conflicting. Part of me wanted to ignore it so I could get it over with sooner. But sharing analysis with a wrong conclusion can be harmful. So I dug deeper. The week turned into months. My conclusions became more nuanced. I was thinking that if I changed my opinion or found an important piece of evidence yesterday, it would be foolish to stop thinking and searching for more evidence today because the probability of me changing my opinion again is significant. I think that this is a useful heuristic. If I would’ve time-capped, I would’ve published the wrong conclusion. Maybe someone would've later did more research and correct it, but that would’ve required more effort spent on the issue overall, overall research on the topic would be more difficult to comprehend because it would be in multiple places, and some people would hold incorrect opinion despite the new research because they only read my incorrect research.
Great post, upvoted!
Nice post Joey. I totally agree with the effectiveness of the Time capping approach. I have read your old post on the process you follow and the Google doc (don't have the link right now) that you use for collaboration and found it very useful for my own work. Thanks for sharing.
I call this "timeboxing", and it's been really useful to me when I can bring myself to do it. I'll also note that Giving What We Ca has acknowledged that they should have spent less time on certain research: