Do declining Hollywood revenues mean fewer movies or smaller mansions?

| Gabriel |

Matt Yglesias has a characteristically smart reply to my post on the effects of piracy in Hollywood vs music. Specifically, he notes that movie stars make a lot of money and we can expect losses to piracy to mostly affect them. That is, much as I argued that music piracy is not that big a deal because it mostly affects rent dissipation in the form of payola, Yglesias makes a parallel argument about elite labor compensation in film. I agree with this argument in part (in fact I anticipated it in my penultimate paragraph) but I think he goes a bit far in expecting that the cost will  primarily be felt in the form of reduced star compensation with little impact on the filmgoing experience.

Before quibbling, I should say that I totally agree that reducing the rewards to elite Hollywood workers will not greatly reduce the quality of our stars. If you think, as I do, that fame is endogenous and only very loosely coupled to native talent, then we might have some retirements in the short-run but in the long-run we can expect very little decline in movie star or screenwriter quality if the top movie stars and screenwriters take a pay cut. The real issue is not that cutting above-the-line costs will hurt quality but that above-the-line costs are a minority of all costs.

Since Yglesias brought up Tom Cruise, let’s use the example of Mission Impossible: Ghost Protocol. IMDb estimates the production budget at $145 million. For the sake of argument, let’s assume that Tom Cruise got paid $20 million, and maybe another $20 million spread between the other above-the-line talent like Brad Bird and Jeremy Renner.  (These are probably overestimates, but I’m trying to make a conservative argument). This still leaves at least $100 million for below-the-line costs. If the expected value of green-lightening MI4 were cut in half by piracy there’s no way this could fall entirely upon Tom Cruise and other elite labor, even if we assume they would work for free, because their total compensation was probably less than a third of the budget.

Making big tent-pole films is really expensive and not just because the director, screenwriters, and starring actors can bid up their compensation. In Hollywood accounting these people are known as “above-the-line” and their wages are, to a first approximation, equal to the expected value of their marginal contribution to the film’s revenues. Although the guilds set minimums, anyone you’ve heard of gets paid much more than union scale and in practice above-the-line wages are negotiated for each elite worker on each film. (This is why people need agents).

In addition to the marquee collaborators, there are lots of craft workers who contribute to films. To a first approximation their wages are set by unions and their cost is a multiple of the number of days on the shooting schedule. Spectacular shots take skilled labor and time. For instance the shooting for MI4‘s amazing hotel climbing sequence alone took eight days, all of them overseas where you have to fly in and put up the expensive union crew. Yes, they could do this more cheaply with digital but I think we’ve all learned that over-reliance on digital matte painting leads to terrible acting performances and a generally creepy uncanny valley feeling.

In much the same way that we can imagine cramming down the compensation of red carpet celebrities we can similarly imagine cutting the wages of below-the-line workers. However this is a much more difficult proposition, even if we put aside the “Hollywood craft workers are the salt of the Earth” talking point that Chris Dodd likes to use to pitch SOPA to left-wing audiences. Key grips are in some sense “overpaid,” but not by nearly so much as are movie stars. The fact is that if you want a master craftsman you need to pay a solid middle class living or that person will defect to another industry. We can imagine cutting celebrity wages by maybe 90% without much hurting labor quality but there’s not nearly as much fat to cut on the below-the-line side, maybe 30% or so. If you want to get below that it’s not going to come out of pay cuts but reduced labor utilization, and by extension lower production values. Contrast the flat lighting in Dawn of the Dead with the rich depths in Land of the Dead  and you’ll get a sense of why the latter film had a budget that was eight times higher in real dollars.

That’s in the long-run. There’s also an issue of what the transition would look like. As I mentioned above, movie star compensation is negotiated ad hoc for every film. If revenues drop appreciably the quote price for above-the-line workers will follow almost immediately. In contrast, getting a 20% or 30% clawback from both craft workers and non-name creative workers is a different procedural ball of wax. These wages are set by union contracts and would have to be renegotiated with SAG, WGA, and IATSE. Paring back the cost structure of an industry under threat is really hard and creates lots of resentment. Moreover the studios are notorious for their sketchy accounting which means that workers will understandably find it difficult to believe “we can’t afford it.” As such an attempt to clawback from union contracts would almost certainly provoke a series of strikes, which not only disrupts shooting during the strike itself but for any film that can’t be completed before a possible strike date.

So I agree that a lot of reduced revenues will be felt in lower star compensation, but a lot of it will cash out in terms of lower compensation for below-the-line workers and lower production values. And that’s after we suffer through a series of strikes to renegotiate the union contracts.

Advertisement

Sympathy for the IP Industry

| Gabriel|

[Cross-posted at NR‘s Agenda blog]

In the arguments over SOPA, I’ve seen a few arguments from people I respect that piracy basically doesn’t matter. These arguments strike me as somewhat plausible but probably wrong and grounded in wishful thinking that a solution being unpleasant means that the problem it addresses is nonexistent. This is not to say that I support SOPA, for I do not. My main intuition on this is that an industry that sponsored the Sonny Bono Copyright Term Extension Act has forfeited its claim to our sympathies. Thus even when it has a legitimate grievance, I am inclined to give it only mild weight. Thus I tentatively favor the Megaupload suit but I’m gonna say “sucks to be you” when the industry demands escalating the fight against piracy into the top priority of US trade diplomacy and a total war waged on the terrain of the internet’s low-level infrastructure. Nonetheless I think it’s important to clarify just how complicated estimating the effects of piracy are.

Much of the debate centers on first-order effects of immediately displaced sales. That is, at a micro-economic level how much does a pirated copy of work i substitute for a legitimate copy of work i. Early on the IP industries had some estimates premised on the idea that each pirated transfer represented a foregone retail sale, which implied the absurd counterfactual that absent piracy we would have seen a massive boom in sales. Critics have appropriately rebutted these studies with the reminder that demand curves slope downwards so quantity demanded at price $0 will be considerably higher than quantity demanded at price $14.99 (at least if we assume low search costs).

I would say that a more important issue is the second-order effects. This appears in some of the anti-anti-piracy arguments as some variation of “people won’t pirate if you make content available in a convenient format at a reasonable price” or “piracy is a customer service issue.” I think this is basically true as an empirical matter and it’s certainly a very parsimonious description of my own behavior. The trick is that you can rephrase the argument as “the threat of piracy has forced distributors to lower their price points and adopt formats that are less desirable to the producer.” As Kernfeld notes in Pop Music Piracy, this is a very old pattern. Basically, producers create some kind of format at a high price point and consumers buy it until a pirate comes along and both undercuts them on price and introduces format innovations. At this point the incumbents try for awhile to suppress it, before giving up by adopting the pirate’s format innovations and dropping their price point. That is, the incumbents ultimately realize that the only way to deal with piracy is a “convenient format at a reasonable price.” Kernfeld emphasizes mid-20th century pirated songbooks as competition for legitimate individual pieces of sheet music but he also applies it to the more familiar case that the music industry only gave in to low price (and eventually DRM free) digital singles to replace high price CDs as a desperate rearguard action against music file-sharing.

Recorded music revenues have dropped precipitously since the late 1990s but only a minority of this was the direct result of sales substituted by piracy. Rather the great bulk of the drop was from the shift from CD albums at a price point of $15 to digital singles at a price point of $1. We have in fact seen a large increase in units shipped, but mostly in digital singles at the low price point. You can see this clearly by looking at Census Statistical Abstract table 1140 and contrasting the unit sales in the top half with the dollar value in the bottom half. To fully make up retail sales we would’ve needed a 15-fold increase in volume and this has not happened. Even if we appreciate that there digital implies lower costs (no inventory) and think about wholesale rather than retail, we’d still need something like a 5-fold increase in sales to make up for lower revenues.

One important consideration of second-order effect is that it means you have to think at the macro rather than the micro and this makes analysis difficult. That is, many researchers have looked at how often a song is pirated and tried to estimate elasticities with legitimate sales. This is a good way to estimate the proximate effects of piracy but it misses the much more substantial second-order effects since the record industry has not dropped its price point only for heavily pirated songs, but for all songs. Let’s assume that lots of people would pirate Eminem but nobody would pirate Norah Jones. If the record industry switches to a digital downloads model to protect sales of Eminem, this will still decrease the dollar value of Jones’s sales. Conversely if higher concert revenue helps make up for declining recorded music revenue this shift applies in nontrivial ways at the micro level.

We also see second-order effect in the film industry. As Reihan has noted (in discussing Doctorow’s summary of Waldfogel and Danaher), one of the ways the movie industry has responded to piracy is by accelerated releases. It would be a mistake to claim that opening wide is driven entirely by piracy. The fact is that it’s a long-term trend since the mid-1970s and also has to do with supply side issues of promotion and a screen glut. However piracy is a part of opening wide, particularly in terms of major releases opening simultaneously world-wide. Dubbing, prints, and promotion on a global scale is extremely expensive and if studios had their druthers they’d rather postpone it until they had an estimate of how well the film does domestically. As you can see by looking at the breadth of release dates, they did in fact drag out foreign releases in the 1970s and 1980s, but in the 1990s and 2000s they’ve been getting increasingly close to a simultaneous world-wide release. This is not so much an issue of foregone sales as it is of increased expenses but it is a way that providing better customer service so as to avoid piracy does cost the industry. That’s not to say that this is conclusive, as Reihan and Doctorow observe it is probably more efficient and just to have Hollywood bear the private cost of accelerated release than to have governments (and private ISP companies) worldwide bear the costs of aggressive enforcement.

Now for the sake of argument let’s take as granted that direct costs of piracy and indirect costs in the form of better “customer service” are costly to the industry and ask what are the consequences for what gets made? We actually have some evidence that this has not much affected the quantity and quality of music but I find it difficult to be as optimistic about film.

The difference has to do with how money gets spent in both industries and in particular appreciating how promotion is the key resource in the entertainment industry. In the music industry the ratio of promotion to production costs is about 10:1. Some of this is about creating fame, but much of it is about allocating fame. This is important because to a first approximation fame is inherently scarce. As Ricardo argued in his analysis of the Corn Laws, when quantity is fixed any change in demand accrues to factor producers. That is, if the sales of pop music decline over the long-run this will cash out as increased consumer surplus and declining value prices for advertising. A very high proportion of music promotion costs does not occur in the general advertising market but in specialized markets, including payola. This means that over the long-run a drop in music revenues will in large part be felt by radio stations and others who specialize in promoting music.

In contrast, the promotion to production ratio in Hollywood is about 1:2. That is, the rule of thumb is that prints and promotion are about half again the cost of getting the film in the can. This means that decreased profits will mostly hit Hollywood itself rather than a related industry. Since stars are residual claimants and below-the-line workers make solid middle class livings, some of the pain will hit in the form of lower labor compensation, however you can’t lower production costs without eventually hurting production values. [Update: Also see Yglesias and my own follow-up on this issue].

More broadly, I think we need to be skeptical of free lunch thinking that if a policy has undesirable consequences this doesn’t mean we have to pretend there is no real problem it is addressing. It’s a common position to say “I don’t like bullying tactics, bad faith arguments, and rent-seeking of the IP industry, therefore piracy is not a problem.” I sympathize with this frustration but it’s more intellectually honest to take seriously that there might be a problem that we decide it is better to leave unsolved.

That depends on the meaning of “all”

| Gabriel |

Public Religion Research Institute has a poll showing denominational breakdown of opinions on the question “All employers should be required to provide their employees with health care plans that cover contraception or birth control at no cost.” They made a very nice bar graph of this that has been widely picked up (eg, Dave Weigel at Slate). The graphic shows that only among white evangelicals do a majority of respondents disagree with the prompt. Most notably the poll puts Catholics against the official position of their church. This is very interesting but it also illustrates a few problems with interpreting poll results.

First, the underlying question is a 5-point Likert scale and the summary graphic lumps together “mostly agree” and “completely agree.” It’s a fool’s errand to try to translate Likert scale responses into equivalent legislation, but it’s plausible to imagine someone putting “mostly agree” whose more elaborated thought would be “in general I support such coverage, but I think there should be a conscience exemption.”

Moreover, we can even imagine someone who would think that if primed but it doesn’t occur to them and so they say “completely agree.” In fact, the poll did prime people in a follow-up question by asking specifically about health plans offered by “churches” and “religiously affiliated colleges and hospitals” and this wording gives you  about a ten point swing.

Now in a way this doesn’t make sense. To a mathematician the set “all employers” includes the set “religiously affiliated colleges and hospitals” but that’s not how people think. Rather people basically read “all” as “typical” or “unmarked” and so they may revise their theoretically absolutist position when you present them with a compelling special case.

You can see this in GSS “violent experiences” module (much of which is actually about opinions), which has long asked when we would approve of ordinary people and cops hitting other people. The GSS first asks “are there any circumstances” (HITOK, POLHITOK) and then asks about several specific scenarios. In the first two waves, NORC treated HITOK and POLHITOK as skip codes. That is, if you said no to “are there any situations” where you’d approve of violence they took you at your word. This was a mistake as revealed in later waves when they started asking the follow-up questions anyway, regardless of responses to the generic questions. It turns out that quite a few people who categorically disapprove of violence in the abstract will approve of it once you prime them with a compelling justification for violence. For instance, about 70% of theoretical pacifists approve of hitting a man who is beating a woman. Actually, some people don’t even need a compelling justification, about 3% of theoretical pacifists approve of hitting a protest marcher and about 8% approve of a man hitting a stumbling drunk who accidentally  bumped into his wife.

Moral of the story, you can’t treat poll responses as things that you can plug into syllogisms. Rather you should ignore words like “all” or “never” and realize that in practice they mean “usually.” For the more compelling case you have to adjust up, for the less compelling case you adjust down, and if at all possible you use a wording that describes the particular scenario rather than inferring from generic conditions. In this respect it’s a bit baffling that PRRI emphasized the generic wording rather than the more specific wording that was pertinent to the bishops’ revolt against the administration. On the other hand, the generic wording is suggestive that an effort to eliminate the mandate entirely is likely to fail badly.