Does the New York Times Adequately Pay Its Crossword Puzzle Constructors?

This article addresses the supposition that the New York Times (NYT) does not adequately pay the constructors who create its daily crossword puzzles. We will provide some information and analysis, but leave it to the reader to decide whether they find the arguments to be compelling.

Let’s begin with some general facts:

1. The daily puzzles printed in the New York Times are created by the general public. Anyone can submit a puzzle to the editor (currently Will Shortz). He culls the submissions, selects which puzzles to use, and works with the constructors to create the finished puzzle.
2. Constructors are paid approximately $450 per puzzle that is used (depending on a few factors, such as whether it’s a Sunday, and how many of their puzzles have previously been used).
3. NYT has approximately 350,000 digital subscribers to the crossword puzzle, who each pay approximately $40/yr. Of course there is some additional income, which is harder to quantify, that the puzzle generates by enticing customers to purchase the print edition of the paper. And in sales of crossword collection books and other merchandising.
4. There are 365 NYT crossword puzzles published each year. (They also provide other daily puzzles that we will mostly ignore for the purpose of this discussion.)

Moving on to the analysis, I have seen two primary arguments for why NYT crossword constructors should be paid a higher fee for each published puzzle:

1. Fundamental fairness requires adequate payment for revenue-generating services
2. Higher fees will result in higher-quality puzzles

 We will now take a look at these arguments individually.

1. The Fundamental Fairness Doctrine

This argument is similar to the concern where corporations are raking in large profits while paying their CEO hundreds of times the salary of the average worker. It just seems wrong to do this, at a basic level of justice and morality. In this case, the argument is not so much that the editor (Shortz) is overpaid, but rather that the newspaper is taking massive crossword profits in their coffers and not returning them to the constructors whose sweat and tears are creating the product.

One flawed argument that you may see in this regard is to estimate the total annual crossword income (e.g. $40 x 350,000 = $14 million), divide it by the number of daily puzzles (365), and compare that per-puzzle revenue figure ($38,356) to the amount that crossword constructors are paid for each puzzle ($450). It is easy to see that $38,000 coming in is a lot more than $450 going out. This argument completely ignores the expenses of running the crossword puzzle operation. Mr. Shortz has multiple employees on staff, including Deb Amlen and Sam Ezersky. Another huge expense must be technology. Digital subscribers have very high expectations with regard to quality of the website where the crosswords are hosted. This includes uptime, user databases (storing past results and statistics), QA testing, and support across multiple platforms including browsers, devices, apps, and so on. The annual IT budget is likely a very significant expense, with a large contract between NYT and the company managing this enterprise. Then of course there must be miscellaneous other expenses such as cyber insurance, advertising, travel, reference materials, NYT office overhead, and maintaining crossword databases, to name a few.

It should also be noted that digital subscribers are getting access to thousands of past puzzles dating all the way back 24+ years to (currently) November 1993. And there is content being developed each day beyond just the daily puzzles, as mentioned earlier. So for many subscribers, not all of their $40 is being spent for the sole intention of doing the singular daily puzzle. And we should keep in mind that it is not generally considered immoral for a business to have some profit centers that make up for other divisions that are loss leaders.

Can we think of other endeavors that rely on individuals who are underpaid? One example that comes to mind is blood donation. Human blood is a billion dollar industry, with its contributors typically receiving no more than a few donuts and maybe a t-shirt. If large sums were paid to blood donors, the unintended consequences would actual harm the blood supply. (People in high-risk groups would be more likely to lie on the questionnaire, for example.) Then there is The Masters golf tournament - another 100 million dollar annual enterprise that employs thousands of volunteers who simply have a love for the course and tournament. Is it possible that people truly love the NYT crossword? Well, surely there are quite a few who would love to see their puzzle published, regardless of the remuneration.

So how much is the net revenue being generated, and how does that compare to the stipends being given to the constructors? Do constructors receive additional payments when their puzzles are sold in compilation books? Only a few people deep inside the NYT building would know the answer to these questions, and they are unlikely to release the budgetary numbers that would help us put this question to rest. But surely it is not just a simple matter of comparing $38,000 to $450.

2. Higher Fees Would Result in Higher Quality Puzzles

This is perhaps the more subjective and tantalizing topic. Assume for the sake of argument that even after factoring in the expenses noted above, NYT could afford to pay out ten times as much for each puzzle ($4,500). Would this result in higher quality puzzles? This is such a subjective question that it almost seems silly to talk about it. But we will try anyhow.

Is there evidence in other fields that higher prices yield higher quality? For a tangible asset that can use better materials, this makes sense. But constructing a crossword seems closer to a discipline like poetry or art. It is hard to imagine a poet who can “flip a switch” and create “better” poems for a bigger fee. Perhaps a steep commission could even have an adverse effect on such a creative endeavor. And when we think about art, many people would likely have a hard time establishing the “quality” of a million dollar piece over a $100 one.

Is there a group of top constructors who are forsaking the NYT puzzle because its paltry fees make it not worth their while? Surely most in the industry would agree that Brendan Emmett Quigley is one of the top constructors today. He has submitted 176 published NYT puzzles, including several in 2017. Patrick Blindauer – 64 puzzles, including several in 2017. Patrick Berry – 227 puzzles, several in 2017. So it would seem unreasonable to say that NYT is not regularly attracting extremely talented constructors. On the other hand, certainly there are some top constructors (Matt Gaffney? Peter Gordon?) who choose not to submit to NYT due to the low payment rates.

The next question is whether doubling or tripling the pay rates would encourage an extra cohort of would-be constructors to get off the fence and pick up their constructing pencils. On the surface, this seems like an expected and desirable result, since presumably more total submissions equals more quality puzzles for Mr. Shortz to select. On the other hand, anecdotal reports indicate that he already receives many more puzzles than he could ever publish. So any marginal increase in submissions could ultimately translate to just larger piles of candidate puzzles stacked up on his desk. And would these extra puzzles coming in turn out to be of a higher or lower average quality than what he was getting before the increase?

There is something else that could happen when you raise constructor fees. Perhaps a group of professionals will begin spending a lot more time generating puzzles for NYT, which could crowd out the amateurs. The puzzle could lose some of that special feeling it has on those frequent days when there is a first-time constructor. This has already started to happen, as new full-time staffer Sam Ezersky has been recently publishing more puzzles than a random contributor (with no disrespect to Mr. Ezersky, an excellent young constructor himself). More generally, there is a romantic notion of an amateur working to create a great puzzle for the love of the craft, or just to see his or her name on that byline, as opposed to a career constructor cranking out puzzles as a daily occupation.

Finally, a personal example. I used to work as a volunteer for a municipal fire department. We knew the town saved millions each year by not paying us (other than a small monthly clothing allowance), but it was something we loved to do. The city leaders determined that a paid firefighting staff was not necessary, as the volunteers met all of the goals of the fire service. They concluded that making a change to paying full time wages would not meaningfully increase the quality of the department. Perhaps, even in our modern capitalist society, there are still some disciplines where money is not the only route to quality.

In summary, could the NYT afford to pay more to its crossword constructors? Very likely yes, but not hundreds of times as much based on a simple calculation. Would paying the constructors more improve the quality of the puzzles? Until such time that they agree to run a double-blind study with varying fee schedules, this is destined to remain a topic of pure speculation.

Looking Ahead

Taking a step back to look beyond the NYT puzzle, crossword puzzle construction does not seem to be a lucrative profession. We do not find it alongside Computer Programmer and Nurse in any lists of top positions for new college graduates. Can anything be done to bring more money into the profession?

One thing about our modern capitalist society is that the quality or value of an activity is not what drives its success; only whether the activity is capable of generating dollars. In an ideal world, you can have government fund those activities that are for the greater good but just don't generate capital (having a taxpayer-funded Poet Laureate, for example). And we do that to some extent. Should we be advocating for government subsidizing of crosswords? That is probably not going to be a popular solution.

The other problem the crossword industry has, and maybe the elephant in the room, is one of scale. If you have a group of 350,000 people, they all need a separate mattress and a separate house, but they can all do exactly the same single crossword puzzle. So if you think about it, how many crossword constructors do we really need? Perhaps the problem isn't that the payouts are too low, but that we just don't need very many constructors.

End of Year Review

I am trying something new starting this year. On the last day of the year, I will look back and list anything new that I did that year for the first time. The goal is to have at least one significant item each year.

Here are my items looking back at 2017:

• Wrote a play and submitted for performance
• Wrote a song and coded it up in Musescore software
• Went to Paris, France
• Had a "clean run" through a crossword tournament
• Went caving
• Canceled cable TV service
• Restored files on a crashed computer from an online backup
• Kayaking on the Potomac River
• White water kayaking
• Played golf with temperature below 30 degrees