Another Cost of Global Warming

This paper documents a small but systematic bias in the patent evaluation system at the United States Patent and Trademark Office (USPTO): external weather variations affect the allowance or rejection of patent applications. I examine 8.8 million reject/allow decisions from 3.5 million patent applications to the USPTO between 2001 and 2014, and find that on unusually warm days patent allowance rates are higher and final rejection rates are lower than on cold days. I also find that on cloudy days, final rejection rates are lower than on clear days. I show that these effects constitute a decision-making bias which exists even after controlling for sorting effects, controlling for applicant-level, application-level, primary class-level, art unit-level, and examiner- level characteristics. The bias even exists after controlling for the quality of the patent applications. While theoretically interesting, I also note that the effect sizes are relatively modest and may not require policy changes from the USPTO. Yet, the results are strong enough to provide a potentially useful instrumental variable for future innovation research.

From a paper by Balázs Kovács, here. Hat tip Kevin Lewis.


But how do himicanes effect patent acceptance rates? Or power poses for that matter?

I'm glad this was the first comment...

I am a firm believer that driving is more dangerous on beautiful days than average ones, because everyone is "la la la, no worries."

Not that AGW necessarily means more beautiful days, which tend to be in the 70s.

Maybe this just wasn't mentioned in the abstract because it's so obvious but, did they control for time of year? For example the seasons could affect when people are likely to take vacations. The holidays are also not evenly distributed in the year.

Good point. The paper is gated so I can't read it, but it's well known that at the end of every two week period (US patent examiners are informally rated by their allowance rate every bi-week) more patents are issued, as well as at the end of every quarter (also a critical rating period) and at the end of every Federal fiscal year (ends in October). Another factor is that perhaps the effect seen by these researcher authors is random, since I'm not sure how the researchers determined the notice of allowance. There's a gap between the time the examiner gives a notice of allowance and the time the patent is formally published, I hope the authors somehow got a hold of the data for the former and not just the latter. The latter data is online, and easy to find, but the former data I don't think is easy to obtain, unless somehow the researchers got permission from the US Patent Office.

Bonus trivia: rumor has it some US senior patent examiners, who have tenure and cannot be easily fired, used to not even examine patent applications but issue them all together at the end of a rating period to reach their numbers. That's one reason there's bogus patents. The European patent office has a two person team which is IMO a better system for examination. An assembly line of experts would be even better. Crowd sourcing even better, and adopting Japanese / German "laid open" (not even examined until litigation, the way US copyright applications are now) for trivial or small improvement type inventions is even better.

I have access to the paper. They do have a dataset with the date of the actual decision, and they are controlling for week of year and day of week, which should address these concerns.

Yep. From the link: "to control for seasonality, I include year, week-of-year, and day-of-week fixed effects."

I was also going to raise this point. Most federal employees take their vacations in the summer. I'm not sure how or why that dynamic would lead to higher approval rates though. Maybe they bring on contractors who were former employees.

I also supervised student research that showed hours of daylight was a significant explanatory variable for suicides. It's possible that diminished daylight hours makes reviewers more likely to deny applications.

Let's suppose that vacations and daylight hours are the ACTUAL best instruments. Does it suffice that other variables, such as temperature, are highly correlated to these, making that ALSO a good instrument. I mean, does an instrument have to have some good theoretical justification?

Hours of daylight, though, are directly linked to season such that they can't be untangled from other seasonal effects.

That's not true of temperature or cloudiness, which vary with respect to everything else. So it seems like their ability to control for potentially related things is better.

Excuse me while I reschedule my tax appeal...

I was going to +1 your comment, but then it started raining.

If I remember right, make it for right after lunch.

is this a joke/troll or a piece of data mining junk science?

There are lots of similar results. They tend to be small and amusing rather than big or scary.

While these sorts of studies are always subject to concern about whether they controlled for confounding well enough, I see no reason for extreme skepticism in this particular case.

There is a very plausible mechanism here: weather affects mood, and mood has an effect on decisions where there is some level of subjectivity.

The key word here is "instrument." The purpose is not to show that temperature is an explanator, but that when you have endogeneity between two theoretically sound independent variables, you can substitute this as an instrument for one of them and avoid biased estimators.

I had a patent rejected because an image of the product appeared on an obscure YouTube video in Malaysia a couple days before the application period.

Must have been a heat wave and coming up on lunchtime.

"A prisoner's chance of parole depends on when the judge hearing the case last took a break, say researchers who have studied decisions in Israeli courts. As judges tire and get hungry, they slip towards the easy option of denying parole, say the researchers."

Absolutely. Never call or make an appointment to meet someone before lunch when a positive outcome is needed. It's much easier to wheedle a person when they've got a full belly and are on the brink of taking a nap. Always save your request for after lunch time.

Do they take into account the fact that many USPTO patent examiners work remotely most of the time? Some far outside commuting distance of Alexandria? USPTO examiners are allowed to work from home full time after the first year.

Great point. But if the researchers did, in fact, have that data, wouldn't that make the results more robust? They wouldn't have Washington DC specific or daily commute factors coming into play.

The US has incredibly sophisticated and effective HVAC systems in almost all residential and business buildings. Do the patent examiners work outside, at picnic tables perhaps?

Did they control for umbrella patents?

Why isn't this a benefit of global warming? More patent allowances means more protection for inventors, which fosters innovation. Cold weather depresses the examiners, who undermine U.S. competitiveness with their negative attitudes; warmer temperatures will lead us into the broad sunlit uplands of higher productivity growth.

i suspect most people on this blog would be of the opinion that these patents undermine innovation, due to the way incumbants exploit them to create barriers to entry for new entrants

I wonder if the effect size is big enough ro use the random variation as a natural experiment into the effectivness of patents

Humorous, but next time you hear someone whining about "We need more tax money to study the effects of global warming" .... this is the kind of crap they are talking about.

This has got to be a parody, right? Okay, a few questions. As someone mentioned, Is the author aware that most patent examiners now work from home? And most modern homes have air conditioning? So if your toddler is screaming, or you have a crummy home office with no window (or a great home office with a view of the park) this is all figured into the mass (or mess) of data? Patent examiners are not sitting in the open air doing their job. They are in enclosures that are light and temperate controlled. You know what? On an overcast day I'd turn on another light. And did they control for the side of the house the examiner works in? Our home office is in deep shade until about 6 PM in the summer. That's because, if the author does not realize it, ambient light into an enclosure depends where the room is facing.


Also, the MR headline here is misleading. Global warming and its connection to local weather events is highly tenuous. You can't even get agreement that hurricanes have anything to do with global warming -- even NOAA isn't willing to commit.

This reminds me of the type of paper that is lampooned at New Real Peer Review on twitter.

Is this a dry and witty parody or more idiotic global warming propaganda? Please clarify.

Good news everyone. Denial trolls are about the least believed source for science. "Americans are most likely to say niche science sources get the facts right." Scientists, not trolls.

Like belief has anything to do with science. Or consensus either; somehow I missed that in all my understanding and practice of scientific method. Good thing Galileo went against consensus and belief.

It would be interesting if they identified which hot days were just run of the mill hot days and which days were a result of global warming and caused patents to be rejected.

I didn't pay $35 to read this, but this would seem to be a potentially great example of the garden of forking paths (aka researcher degrees of freedom).

People who breathe eventually die.

I am a patent examiner. I have access to the original paper. The public database that the author used shows the date at which the examiner completed the office action, which the author assumed to coincide with the day the decision was made. This is not a good assumption. The author could not see the time of day an office action is submitted, but I have that information for my own cases. The first (and only) rejection I looked up was submitted at 9:30 AM which means that I probably made the decision of how to handle the case on the previous day.

Well that makes sense. Of course! The only value to the nonsense "peer reviewed" science paper is to demonstrate how ridiculous science has become and how people jump on silly papers to religiously support global warming. The REAL CRISIS is this: - the scientific method is not being practiced and science has become a broken institution.

Comments for this post are closed