Le cinéma et l’argent, Nathan, 1999.

Publication edited by Laurent Creton.

Paris set for Les Amants du Pont Neuf

An unusual event took place in the autumn of 1991. A film came out, Les Amants du Pont-Neuf, and everyone knew that it was one of the most expensive films, if not the most expensive, in the history of French cinema: 130 million francs! But the extraordinary thing wasn’t that. The most troubling element about Léos Carax’s film was that the major part of the money spent couldn’t be seen on screen. A series of unfortunate coincidences, reshoots in new sets and a somewhat careless management had contributed to raise the budget to over 100 million francs for a film that, on screen, appears to have cost 30, after having initially been planned for an artisanal shoot in 16mm. Everyone, the industry and the audience, was aware of this incongruity. But it was what was to attract the attention of viewers and help the film have a brilliant run, which would’ve been very satisfactory had it cost 30 million, but was catastrophic for a work of 130. An astounding number of articles were dedicated to events surrounding the film before its release. A striking contrast with the reviews of the film themselves, which didn’t exceed standard length and even suffered from the exclusivity granted to the analysis of the shoot. Everyone was dying to see the film. And I remember very well that I went to the first show of the film burning with impatience.

For most people, the interest wasn’t as much in going to see a good film as in finding out how one could spend so much money, where it all went, and in gazing at the monster. You went to Les Amants du Pont-Neuf like you read the delectable annual report of the accounting office that denounces various kinds of waste in the administration’s spending. Most viewers were aware that their desire to know that couldn’t be fulfilled since the money wasn’t to be seen in the result.

The colossal publicity for made for these Amants—there was even a film about the film—cost nothing: journalists thronged to get more information, so there was no need for a flashy campaign or scores of advertising billboards. The promotion cost for a film that enjoyed two hundred and sixty thousand admission in Paris region was rather small: it was as small as the shooting budget was extravagant. You could even wonder if the entire affair wasn’t a brilliant bluff, if the budget wasn’t disproportionately blown up just in order to get all this free publicity. Unfortunately, it wasn’t. There’ve been such cases before: every evening of the shoot of Foolish Wives (Stroheim, 1921), Universal put up fictional numbers purporting to show the ongoing cost of the film’s production at Times Square; or the wily Russell Birdwell, PR agent for Alamo made by John Wayne in 1959, who publicised a highly exaggerated total cost in order to garner the sympathy of exhibitors, who extended the film’s run in their theatres, and in order to increase the interest of voters in the Oscar race.

A pure concept

We can cite two examples which come close to Les Amants du Pont-Neuf, but producers are careful not to acknowledge that the money can’t be seen in the result, even if the observer wonders where it could all have gone. The trick lies in making sure that no one has the idea or the cheek to raise the question of wastage. The difference arises from the fact that, in Les Amants, the wastage took place during the shoot, involuntarily, while, in the films we are about to study, it was in some ways anticipated even before the first day of shoot.

Jean de Florette (Berri, 1984) was produced in a budget of 59 million francs (of which 13 were above-the-line costs for star actors and auteurs, which is not excessive). So 46 million was spent on the rest, whereas we “see” 15 or 20 at most on screen: only five shooting locations – Jean de Florette’s house, that of Papet, the two villages, the scrubland – in natural settings within a radius of 500 kilometres, five principal actors, some secondary roles, some extras, no chase, no stunt, no short edits, no special effects (except one, a successful one at that), no sumptuous period costumes. The impression of poverty the film gives goes hand in hand with the material poverty of the protagonists. We really get the impression that if so much money was spent (although we can’t totally rule out a bluff, but I think it’s improbable) it was for the filmmakers to convince themselves that they haven’t left out anything, to say that they have spent a lot of money to impress the gallery, coproducers, distributors, exhibitors and the public.

More ambiguous is the case of Le Garçu (1995), which declares a budget of 67 million (of which 15 above-the-line for big names in the credits), even more expensive than the big spectacle of Captain Conan or Ridicule. I must say I almost died of laughter looking at the cost estimate of Garçu, since in no way does the result allow us to imagine a total cost of 50 million. I’m all the more comfortable saying that because the film is, in my opinion, a real masterpiece. It’s an intimate chronicle revolving around a few characters, featuring only one really popular actor at the box-office. Considering the fact that Pialat shoots a lot of takes and sometimes reshoots a part of his film afterwards, we could estimate the visible cost of Garçu at a maximum of 15 million. We can doubt the veracity of the declared budget here: the astounding figure of 7 million for copyright (music and royalties, while the script and dialogues were mostly improvised), or the equivalent of the total cost of two or three Rohmers!

We can notice a similarity with Jean de Florette: both films feature Gérard Depardieu, who appears in the cost estimate against an amount that’s modest for his reputation. Does this mean that Depardieu only accepts to participate in films with a very high budget, official or real? Or that, as soon as Depardieu comes on board, producers manage to increase costs so as to implicate all economic collaborators a little more, to have fun or to even inflate their contribution to the virtual general expenses? Perhaps it was also, in this particular case, a way to prepare for the overspending that Pialat was accustomed to. Difficult to say.

These three different cases illustrate the phenomenon of stated and not-apparent money. Observers take the classification of expensive films at face value without questioning it. Money is a purely abstract concept and nothing else.

This principle can also be applied in reverse: films that cost very less, but seem luxurious. This is how, thanks to its lighting and constant innovation, The Blind Owl (Raoul Ruiz, 1987), which must’ve cost about 5 million francs, appears infinitely richer than Jean de Florette, despite its budget of 59 million. A costume drama like Let Joy Reign Supreme (Bertrand Tavernier) could be completed in 1975 for 4 million: a miracle that we are hard put to explain. The five hours of Jacques Rivette’s Jeanne d’Arc (1993), for little more than 20 million… movies made in developing countries could boast of an absolutely phenomenal quality-price ratio (Farewell My Concubine, Antonio Das Mortes, The Holy Mountain, Oliveira’s No).

Or, simply, the set decorator’s shrewdness enables some incredible savings: we know that, for Reign of Terror (Anthony Mann, 1949), the brilliant Cameron Menzies had come up with a prism system that could multiply the number of extras seen on screen manifold.

We’re dealing here with an international phenomenon in both ways, in profligacy as well as in parsimony: we know, for instance, that many American productions for major companies have their budgets blown up by their distributors/financiers, thanks to the inclusion of imaginary or useless general expenses. On the other hand, the scenarios we are about to examine below seem to be typically or exclusively French.


When we have the means to turn over the cards of this strange game, it appears that the official budget of French films, as they figure in the Centre du Cinéma magazine (CNC info) differ markedly from reality, at least as far as modestly-budgeted films are concerned. A precise report written by a student of Paris III estimates, after an interview with the director, the cost price of Inner City (Richet, 1994) at 430,000 francs. The number quoted by CNC info is 2,010,000 francs, 469% of the real cost. Two years after the answer print of one of my films, my producer gave me a final summary of expenses: 36,000 francs. However, the CNC quotation reached 250,000 francs, 694% of the real cost. And yet, I can’t boast of having set a record.

A Jean Rouch film produced by Pierre Braunberger around 1962 is said to have crossed the sound barrier, with more than 1000%. One only needs to add a zero… In general, we see differences that are less stupefying. Let’s say that, on an average, you must divide the stated cost by two to get the real cost. I will cite some personal examples: 210,000 in place of 410,000, or 110,000 in place of 310,000, or 2,200,000 substituting 4,400,000, or 450,000 in place of 1,191,000. And my case is not at all particular. A recent Italian co-production, The Second Time, is declared at 21 million, while, upon viewing, we’d peg it rather at 8 million. The 11 million of The Phantom Heart (Garrel, 1995) are to be reduced probably to 5. These overestimations are sometimes compensated by underestimations that afflict the most expensive films.

These subterfuges aren’t of the same order at all as those evoked at the beginning of this chapter, where the differences between stated numbers and visible spending were voluntary on the part of the producer, no matter that the film was too ostentatious or this ostentation was simulated. In this new category, the differences are neither desired nor taken upon by the producer. They are imposed on him by the Administration. In France, the production of a film is indeed dependent on authorization, the CNC giving its approval only if the film seems expensive enough to be seen through to completion, that is to say more expensive than the real cost of the film, even if the gap has tended to narrow since the middle of the 1990s.

Why this perpetual hiatus? Between 1947 and 1959, the CNC was used to expensive studio productions. It could never accept the drop in expenses enabled by successful films of the Nouvelle Vague from 1958 on and which was made possible by, among other things, location shooting, decrease in number of technicians and the potential reduction of their salaries1. The technicians’ union, which oversees the committee in charge of dispensing production authorizations, tended to oppose films in which salaries were lower than the minimum professional wage fixed by it and whose number, in its view, was grossly reduced (minimum salary which, let it be said, wasn’t obligatory at all for non-unionized technicians and producers).

In conclusion, filmmakers were better off lying and declaring bloated salary numbers and sufficiently high fees before shooting. Failing to comply, a producer saw his file adjourned or rejected. The shoot, for which everything was carefully prepared, found itself pushed, with concomitant postponement fees (already-signed contracts and engagements) and logistical problems (rescheduling of a shoot that had to take place in a particular season to the following year). Subterfuge was a good tactic: it solved all problems and saved time in the dealings with the Administration.

This taste for overspending, rather pronounced among administrative personnel, stems not only from a nostalgia for the studio era. Government officers, just like politicians who are supposed to head them and for whom they are mistaken, love to see investments increasing. For them, anything that increases is good, anything that reduces is distressing. A director of production at the CNC, noticing that I had made my first feature film for very little money, 50,000 francs, told me: “Okay, I’ll let it pass this time, but I hope that you’ll make a more ambitious film next time…” “More ambitious” meant “more expensive”. I tactically refrained from contradicting him, but I said to myself: “What a moron!” For him, ambition meant spending more, whereas my film, which questioned the inanity of university teaching before May 1968, was one of the most ambitious of the year (and too ambitious, in my opinion). For these officers, victims of a bad education, the best of the best meant always more, spending more, earning more. A very dangerous principle of perpetual ascent that evokes the Tyrolian game picked up by game shows, and which leads straight to a breaking point, to the Tarpeian Rock, to Tex Avery’s King-Size Canary, where the canary becomes bigger than the earth.

At first, I naively proposed authentic cost estimates to the CNC, but the personnel at the Centre seemed alarmed by it. The best officer this organization ever had begged me to make an effort to blow up the estimate a little: “My higher-ups will laugh at my face if I hand them such a poor budget.” Seeing him distraught, I told him after some hesitation that I accepted his proposition. He then started wiping the sweat off his forehead, and I think he was grateful to me for my cooperation. This incident proves that his sense of reality was very diminished. He was the best officer and yet he lived in the clouds. His more strait-laced colleagues lived on the moon. Later, when I asked this good man what minimum estimate I should quote to the CNC for a feature-length project (“800,000 maybe?”), he agreed to accept a budget of 1 million, throwing his hands in the air: “I wonder how you can make a film that is to be shot in three continents under 1 million.” Well, the final cost of the film was 298,000 francs…

To be sure, almost all cinema professionals know well that cost estimates are fudged. A little internal machination that bothers no one… The problem is that a lot of them don’t know to what degree. Cinema officers are in fact the children of Mao, since communist China was reputed for its false statistics…The negative consequence of this system is that very official people, observers, international publications and five-year plans gulp down all the number-backed fabrications of the system without batting an eyelid and build castles in the sky with them2. The paradox is that the falser the numbers are, the more their disclosure swells. The officials fudge these misleading numbers a little more and firmly defend them since truth would cause trouble. Contrary to expectations, certain statistics are even more subjective than critical opinion.

There are other factors that make bluffing inevitable: outside of rare exceptions, the CNC doesn’t give its approval for shooting if the advance granted reaches 50% of the cost estimate, or if the share of a French television exceeds this percentage (it will then be a telefilm), or if the producer’s contribution is less than 15% of the total (a legal obligation that was recently removed). Now, a rather in-vogue producer told me recently that no producer (except Seydoux, Berri and Fechner) invested money in a film outside of sometimes fictional general expenses and overhead risks; that shows how illusory this 15% is. Ultimately, even if the real estimate turns out to be acceptable by itself for the CNC, you must compulsorily “inflate” the expenses in accordance to the abovementioned percentages and especially – this is the greatest disadvantage of the system – “justify” a fictional investment.

The other disadvantage of the system is that, to increase the apparent cost of his film, the producer’s first (and easiest) approach consists of considerably increasing the royalty cost which, filed in public register, will initiate a drive on behalf of the Artists’ Welfare Office, eager to collect subscriptions based on this fictional amount.

This cover-up isn’t without its advantages for the producer (even if, in general, bluffing is not justified by money-mindedness at all). If the producer’s percentage on the revenue was to be limited according to his real investment, it would hardly reach the 7% that is the current norm for general expenses. The producer’s motivation to get good distribution for his film would then be very limited and would hence incite him into passivity, while with bluffing and his own fictional investments, he can get percentages going from 30% to 100% (considering that certain collaborators are not paid from the first franc of the revenue onwards).

This constant, playful inflation of investments is at once an exterior sign of health and a good reason for asking for aid and funds from the State and from everybody: these sums are so disproportionate that poor producers are hard put to make ends meet and more subsidies, more tax reliefs must be offered for cinema. Paradoxically, we could come to consider that the more statistics are fudged, the better films are (and vice versa). The forced inflation means that they are financed in unorthodox ways, which implies an originality lacking in most films.

Contrary to popular belief, these financial subterfuges aren’t just limited to production numbers. To a lesser degree, differences exist at the exhibition stage. This is how, in Film français, my Brigitte and Brigitte was declared as having had 22,155 admissions during its limited release in Paris. The data processed by the CNC, more exact, gives a figure of 19,357 admissions. It’s my distributor who inflated the numbers communicated to the press in order to help the film, to make provincial exhibitors believe in its money-making potential, and especially to cross the symbolic threshold of twenty thousand admissions. A good film like Muriel – ninety thousand admissions in Film français – probably had less than 50,000. It seems that this cover-up job isn’t possible anymore today, everyone having the right now to verify the reality of these numbers, but I hardly believe that: who’d have the time and the interest to verify this? The only difference is that overestimation is disallowed for champions of the box office, where the bluff would be too obvious.

There is always the possibility – rarely harnessed today, but once common among producers and distributors – to buy tickets to their own films upon their release. Film français won’t lie about the numbers, it’s the numbers themselves that will lie. I remember, for example, that the producer of Robert Bresson’s A Man Escaped had bought dozens of tickets in 1956 so that the film crosses the threshold contractually necessary for getting a third-week run3.

We can also mention the fraudulent practices in theatres (and which generally works the other way around, towards an underestimation of revenue, but not always), with double sale of the same ticket, the attribution (in multiplexes) of a theatre to a film that doesn’t belong there when the multiplex manager wants to favour a programme in which he gets a better percentage4.

It’s frequently said that the share of the producer/distributor represents about 40% of the revenue, and can’t legally cross 50%, but that’s to forget that distributors are paid minimum guarantees: this is how my The Comedy of Work made 63 francs as revenue for Auchel, and the distributor received 1,000 francs, 1,547% of the revenue. Am I finally going to make it to the Guinness records?

Snowball effect

A fundamental principle of cinema is the snowball that ends up causing avalanches. It’s for this reason that there’s an interest in inflating the revenue numbers of the first run. Some provincial exhibitors reject a film if it hasn’t crossed a certain threshold in Paris. Even I went to see Diva and In girum imus nocte et consumimir igni because their limited release lasted long (at the producer’s cost); this permanence intrigued me.

Distributors and producers seem to be sensitive to the logic of rivalry. They’re all the more interested in a film when they notice that a competitor is interested in it. One day, at a film festival, a distributor told me that he liked my film a lot, but he couldn’t unfortunately take it since he had no more theatres at his disposition. Six months later, I paid for a private screening of the film at a theatre which turned out to have been controlled by this distributor. At this screening was another distributor who enthusiastically made me a hard proposition. Some hours later, the first distributor called me, outraged: “What? You invited a competitor into my theatre, and she wants to buy your film, while it was I who made the first offer?!” I retorted that he’d told me he didn’t have any more theatres and that he’d refused the film. He answered that all of Paris new that he had three theatres and that he’d offer me a sum greater than the one proposed by the competitor, with the promise of distributing another one of my undistributed films.

Thanks to this experience, I realized that in order to get offers from an economic operator and make it go through the roofs, you absolutely had to sustain the interest of a (real or fictional) competitor, to make yourself seen by him, even at the risk of paying this competitor for his temporary service. This is the principle of the “accomplice” to hawkers. “It always rains where it’s wet”, my peer Jean-Danier Simon used to rightly say.

When my first film, Brigette and Brigette, was presented at the Cannes Festival, in a small private theatre, a distributor called me to the smoking room at the end of three minutes: he had noticed that some of his peers were laughing out loud in the hall, and he wanted to be the first to make me an offer, which was soon finalized. He saw the film only six months later, during its release, and he was much less impressed… On the economic front, the problem is to get the first mark of interest, to engage someone in the film’s cycle. The financier gives his money only to someone who already has, or is suspected to soon have, such a person.

When he started out, Claude Lelouch found it very hard to sell his films. He recounts how he had successfully convinced a German client to buy Une fille et des fusils (perhaps his best film) not because of its highlights but because the distributor had seen him driving around in Mercedes… I also remember that a Venezuelan client had asked me, as a precondition to buy my film, some data about the film’s performance in other countries. I made a very precise list of completely imaginary positive reception in faraway cities, where any verification was impossible: Seoul, Oslo, Nairobi etc. And my film was sold this way to Venezuela.

Similarly, when a foreign client enters a production house with posh offices, he won’t dare proposing a lower price for the purchase of a film. We can cite an amusing experiment in this regard: for his client meetings, a wily fellow had the idea of renting producers, magnificently-equipped offices on Champs-Élysées with secretaries and even a name plate on the door for one hour. And it worked…

In the same order of ideas, we’ve often wondered why American films have had such a clear edge in France in recent years. There are several reasons including this one: earlier, in France, a high-profile film was issued at 50 prints at most. Today, 600 prints are needed given that the film should be shown in as many theatres as possible on the first day. A release today involves a lot more issuing cost than before. As a result, sure-shot products are favoured, those that have already proven themselves and made a lot of money (so mostly American films), which are more reliable in principle than resorting to successful French stars (there’s no guarantee that their next appearance will be popular). As a consequence, grand launches, which are partly responsible for record revenues, will lean towards American bestsellers and professionals will do their best to forge a convergence, even a similarity, of tastes between the two continents. The phenomenon is relatively new, if we consider that the champion of American box-office in 1959, Auntie Mame, wasn’t even dubbed and ran for only three weeks in just one theatre in Paris.

Another variant: to be well distributed, you must engage the most amount of people holding economic powers, distributors, investors and exhibitors, such that they seek to recover their investment by all means. If you make them run the risk of losing money, owing to the extent of their investment, you are sure to get a good distribution. If, on the other hand, you make a film all alone, without involving anyone, you are necessarily at a disadvantage compared to those who have compromised their clients who, in turn, will seek to recover their principal sum most of all. A suicidal reflex at times: the film where a lot has been invested could turn out to be less commercial than one made without a single penny.

As a corollary, it seems that it’s beneficial to intimidate economic operators (without necessarily asking them to invest) by making them believe that, if your film doesn’t work, it’s the entirety of French cinema that’ll be in crisis. They then find themselves invested in a mission. By agreeing to take over from the weak-hearted producer of Les Amants du Pont Neuf, Christian Fechner knew well that he was going to lose money. But he became the film’s saviour, the saviour of the most expensive film in French cinema history. It’s better to spend the most possible amount of money (or pretend to) and appeal to public aid, something the producer of a film made for 3 million francs can’t do.

We can lose sight of an essential principle: what counts the most in a publicly-traded company is the dividend on the stock, and not the turnover. It’s the opposite in cinema. Three million admissions for The Messenger: The Story of Joan of Arc (1999), which cost 350 million francs, a loss in fact. The Horseman on the Roof had two million five hundred thousand admissions and cost 170 million, which comes to 68 francs per viewer. In comparison, The Tree, the Mayor and the Mediatheque, with a cost of one million, had a hundred and eighty thousand admissions, 6 francs per viewer. The return on Rohmer’s film is hence eleven times higher than that of The Horseman and eighteen times higher than that of Joan of Arc. Even so, it’s The Horseman or Joan of Arc which is a model “showcase film”, in total contravention with the “laws” of the economy.

We notice that French film economy clearly follows the American model, but with a lag of several years. Earlier, owing to prudishness, the French economic system concealed everything related to money: before 1947, it was impossible to know film budgets and revenues. Mentioning these amounted to an infringement of business secrets. Only some rare, favourable numbers (true or false) made it to La Cinématographie française before 1949. Now, in the footsteps of puritanical America where everyone has his cards on the table, unashamed to reveal how much he made every year, all these more or less true numbers appear in the press, at the exhibition stage (since 1949) as well as the production stage (since 1978). Like the Americans of yesteryear, we boast today about spending or making the most amount of money possible.

This phenomenon takes place in France at the very moment when America starts to evolve: today, thanks notably to the low level of social security and the return to black and white, the Anglo-Saxons are proud to reveal that their masterpieces – Go Fish, Henry, Clerks, She’s Gotta Have It, Unbelievable Truth, The Blair Witch Project – were made at prices that defy all competition (between 80,000 and 400,000 dollars), unthinkable in France. I’m even tempted to say that they’re underquoting to create interest and buzz in the media. In France, the honest declaration of small budgets continues to give the impression of a lack of seriousness (even though it’s more difficult to make a film with little money than with a lot) and turns against the films. France’s eternal lag over America…


1In this regard, refer to the analysis by Michel Marie in La Nouvelle Vague: une école artistique, Paris, Nathan, collection “128”, 1998

2When the technicians’ union protested in 1973 against the insufficient portion reserved for its members in budgets, the CNC encouraged producers to increase this portion in their agreements: from 1973 to 1974, this went, on paper, from 12 to 21% without a real consequence. We could also mention that these statistics on French cinema include a James Bond movie such as Moonraker, films never started such as Moi, je or L’ailleurs immédiat while excluding Éric Rohmer’s The Tree, the Mayor and the Mediatheque or Robert Guédiguian’s Marius and Jeanette.

3Today, there are friends of distributors who, with their UGC card, continuously punch in at various cinemas to “increase” the number of viewers.

4Exhibitors often base their forecasts about a film’s run on the commercial performance of the director’s and the star’s preceding film. That why, for example, Jacques Doillon found it very hard to make a well-performing film, The Crying Woman, continue its run in a theatre. This success, unexpected with respect to Jacques Doillon’s and Dominique Laffin’s previous performances, was likely to delay the arrival of bestsellers, already contracted for a particular date and their minimum guarantee already paid. The exhibitor at a multiplex moved heaven and earth in order to discourage Doillon’s viewers (reducing the number of posters, hiding stills, turning off the neon lights etc.)


[From Luc Moullet’s Piges choisies (2009, Capricci). See Table of Contents]