Broadway Real Estate and Economic Complexity in the Theater

ASTR 2016 Working Session

Derek Miller
November 4, 2016

This paper examines Broadway from the perspective of theater owners. Those owners today are primarily three private businesses: the Shubert, Nederlander, and Jujamcyn organizations. As with so many businesses, what they actually own and how they own it, how (or whether) they make money, and how they make business decisions, remain essentially unknown. But even if we cannot see precisely how these firms (and their predecessors and smaller-sized peers) operate, perhaps by looking at the environment in which they operate we might extrapolate the thinking that organizes their behavior.

Or so I had hoped. When I first proposed this paper, I set out to analyze one particular problem for theater owners: how they avoid the dreaded “dark days,” days when their expensive Manhattan real estate sits empty and unoccupied. I had thought to examine a dataset of Broadway production dates and weekly grosses to discover when producers kick out under-performing shows and when they grant those shows leeway, and to grasp under what circumstances producers build or sell theaters or convert extant properties from legitimate playhouses to film/television/radio houses and vice versa. In short, I wanted to investigate how efficiently the industry manages its supply of theaters. The answers to such questions have proved more elusive than I had imagined. My attempts to find answers, however, illustrate some essential problems in studying theatrical real estate on Broadway and elsewhere.


In the first case, I assumed that theater owners treated tenants based in part on their weekly grosses. Specifically, I knew that Broadway theater rental licenses include a “stop clause,” a box office gross value below which either party can end the contract without penalty.1 I hoped that the stop clause would partially predict the end of a show’s run, with theater owners dumping poorly-performing tenants for potentially more profitable successors. Visits to the Shubert Archive and the New York Public Library for the Performing Arts turned up an assortment of contracts and their stop clauses. For example, the contract for Neil Simon’s Star-Spangled Girl at the Plymouth Theatre in 1966 stipulated that “Landlord or Tenant shall have the right to terminate this lease if the gross receipts, as defined herein, for two successive weeks shall be less than $22,000 each week.”2 That amount is 40% of the $55,000 the theater could take in weekly. A 1983 contract with Jujamcyn for a production of All’s Well That Ends Well at the Martin Beck Theatre lists a stop clause of $190,000 for a house with an estimated gross of $375,000, amounting to a stop clause of 50%.3 Given those two data points (and a few others like them), forty-five percent of the possible gross seems an acceptable estimate for stop clauses if we wish to extrapolate them over a longer time series.

Unfortunately, theater owners seem to have enforced stop clauses rarely, if ever. At the Royale Theatre in the 1950s, for instance, most shows grossed below 45% of the estimated maximum possible gross without being evicted (Figure 1).4

Figure 1. Grosses and Estimated Stop Clause at the Royale
What explains this behavior? Consider The Star-Spangled Girl’s actual weekly grosses as reported in Variety (Figure 2). As we saw with the shows at the Royale in the 1950s, the Shuberts allowed Simon’s play to run six weeks longer than the stop clause required.
Figure 2. Weekly Grosses for The Star-Spangled Girl
On each of the weeks marked with a red square, the show grossed below its stop clause (indicated by the dotted horizontal line) of $22,000. Why allow the run to continue? The short answer is: because a theater owner prefers making some rent at all times to making no rent at any time. Star-Spangled Girl’s grosses only dipped below the stop clause during the summer, when few shows open. Given the choice between an empty theater and some income, the Shuberts happily let the show by the popular playwright continue past its financial expiration date.

And that income is more substantial than one might think. For The Star-Spangled Girl, the Shuberts were guaranteed a minimum $3100 a week in rent, marked by the dotted line in Figure 3.

Figure 3. Weekly Theater Rental for The Star-Spangled Girl
Even at the stop-clause gross of $22,000 the Shuberts’ take was $4100, a full third higher than that minimum. The stop clause, in other words, provided a lot of buffer above the Shuberts’ minimum rent and would hardly have been worth their exercising in a slow month. Further, productions cover a substantial amount of a theater’s other costs. The Star-Spangled Girl, for instance, paid for: air conditioning at $80 per performance; 73% of the New York City gross receipts tax; $25 per week for supervising tickets; and $175 for the use of a Shubert box office in Macy’s. When the show finally closed on August 6, 1967, the theater sat vacant—and costing the Shuberts money—until the end of September. Even at the minimum rent, the loss over six dark weeks at the Plymouth was at least $18,000, plus difficult-to-calculate costs.

Stop clauses thus offer a poor guide to theater owners’ behavior. Rather, a show’s producers are usually the party most desperate to take advantage of a stop clause. As a recent book about theater production explains, producers should set a stop clause slightly below the weekly break-even point so that they can break their contract quickly if they start losing money.5 Setting the stop clause around the break-even level also ensures that they can keep a theater as long as they are covering their costs.6 In other words, the stop clause primarily protects the producer from eviction in favor of an enticing new show while allowing a quick exit whenever expenses begin to outweigh income. Although I had hoped stop clauses would reveal an important aspect of theater owners’ thinking, stop clauses turn out to signl producers’ behavior. The real imperative for theater owners is always to keep the house occupied rather than empty.

Demand for Theaters

Yet surely there must be cases when theater owners wish to evict—or at least transfer—poorly performing shows so as to make room for a more finanically promising production. When, as a general proposition, might theater owners be particularly aggressive with respect to lingering tenants? When they can drive better bargains; that is, when demand for theaters is high. If the Shuberts could negotiate, say, a high percentage of the gross on a show with an experienced star, they might forego three more weeks of a tired show that is earning around its stop clause. How, then, might we glimpse the demand for theaters? We only know what plays were actually produced on Broadway, not what plays might have been produced had there been greater capacity to house them.

One way to estimate the demand for theaters is to look at the supply of theaters. Under standard economic models, increased demand for a good should drive up the price, thus encouraging the suppliers to increase the supply. In other words, if there are more would-be Broadway plays than theaters to play them in, theater owners should get better deals and, eager to make more money, build or convert more theaters. Let us examine then the number of theatrical venues in use each season over the past century (Figure 4).

Figure 4. Number of Venues Used Per Season, 1915–2014
A few clear stories emerge from this chart. First, theater construction booms during and immediately after World War I, with the number of theaters growing from around 50 at the war’s end to a maximum of 78 in 1926. Then comes a swift contraction, as that number drops in half by the eve of the next war. The period from 1915 to 1940 is thus a huge bubble in theatrical real estate, under which the stock of theaters takes a decade to double and then fifteen years to readjust back to its previous level.

The next fifteen years, which include World War II and the post-war economic expansion, were stable for the Broadway theater market at around 39 venues. That number briefly dips around 1962, bottoming out at 31 houses. Some of this trough results from owners’ renting theaters to television and radio studios or as movie houses. Theaters return to service through the 1970s and the number climbs back to the post-war level and, from 1978 to 1981, to slightly above that level. This climb tracks partly the reconversion of theaters back to legitimate houses (the Bijou Theatre; the Ritz; the Rialto). But it also marks the building of entirely new theaters including the Uris (now the Gershwin), Circle in the Square, and the Minskoff, all opened in the early 1970s. That peak, however, falls off again through the early 1990s—starting with an infamous triple-demolition in 1982—before creeping back to around 40 houses per year, where it remains to the present day.7 Significantly, those changes do not seem to be related to the number of productions each season. Plotting the number of venues against the number of productions each season demonstrates that productions (the red line) followed a relatively continuous post-war decline through the early 1990s before stabilizing (Figure 5).

Figure 5. Venues and Productions Per Season, 1915–2014
Supply and demand in the 1970s and 1980s thus do not seem as closely tied as they had been in previous eras, as when the boom-bust cycle during the Roaring Twenties and the Great Depression matched well number of venues and number of productions.

Perhaps we can make sense of the changing supply of theaters by examining not simply the number of productions but rather the theater occupancy rate.8 More productions might indicate increased demand, but higher occupancy rates (fewer dark days) also indicates a diminished supply of theaters. Figure 6 plots the percent of venues in use on the first of each month on the left axis, and the number of venues used that season on the right axis. (The percent of venues in use equals the number of venues with a show currently running divided by the total number of venues that house at least one show that season.)

Figure 6. Percent of Venues in Use and Number of Venues, 1915–2014

Unfortunately, this chart makes theater owners’ behavior even less comprehensible. Notice the sharp decline in theater occupancy rates around 1970, bafflingly accompanied by a growth in the number of theaters. Why, if venues are more likely to be dark, would owners increase the stock of available theaters? Samuel L. Leiter, in his book on New York theater in the decade, observes that the building of three new theaters in 1972 and 1973 converted “a shortage of theatres” into a “glut.”9 But what Leiter describes as “booking jams” eliminated any sense of glut, at least through the late part of the decade. We can see Leiter’s booking jams in the unusually narrow seasonal variation in occupancy rates during the mid-1970s. Typically, occupancy rates are lower in the summer and higher in the winter, a pattern visible in the high-amplitude, sinusoidal pattern across the chart. But during the late 1970s and early 1980s, the amplitude of the seasonal variations decreases dramatically. In other words, even as the stock of theaters grew and then receded, overall occupancy rates held far steadier over the course of each season than had traditionally been the case. (The large seasonal variation is back in evidence today.) I do not yet know what explains these changes, but the forces involved are clearly too complex to be answered only with data. It is possible that, although the number of Broadway productions did not vary much, more would-be productions arose in the 1970s. Additionally, Michael Riedel hints at a change in the Shubert’s standard contracts around this time that might have encouraged demand: the Shuberts downsized their traditional cut of weekly grosses from 30% to 10%, in exchange for shifting all theater operating costs to the production.10 If and how this would stabilize occupancy rates is not yet clear to me.

The one decision that changed the number of theaters about which we have good information, the razing of the Bijou, Morosco, and Helen Hayes Theatres in 1982, drives home how insufficient economic modeling can be in predicting industry behavior on this scale. Though vehemently protested by many theater personnel, the Marriot Marquis hotel that replaced the theaters drew support from the League of New York Theatre Producers, which hoped the building would be a “powerful economic force for the redevelopment of the Broadway area.”11 The Bijou, Morosco, and Helen Hayes were not demolished because of an excess supply of theaters, nor would increased demand have saved them, particularly given that the whole process took over a decade (1972–1982) from conception to realization.

One further problem arises when thinking about the stock of Broadway theaters: what counts as a Broadway house? While the Broadway League today polices its membership carefully, the historical definition of Broadway is far more nebulous. Two major definitions are important now: (1) the theater has a Broadway contract with Actors’ Equity Assocation, and (2) the theater is eligible for the Tony Awards. But both of those definitions are flexible and have changed over time. Mary C. Henderson, in her influential study of New York theatrical venues, discusses only houses in the Theater District, “located in the streets radiating from Times Square from Forty-first to Fifty-third east and west of Broadway.”12 That excludes venues such as the Vivian Beaumont at Lincoln Center, which has been part of Tony Award calculations since at least 1973 (David Rabe’s Boom Boom Room received three nominations) and was built in 1965. Other problematic houses in the 1970s include the Eden (formerly the Phoenix), which was the first home to major Broadway hits of the decade including Oh! Calcutta! and Grease, and the Playhouse Theatre, where productions garnered multiple Tony nominations.13 In Henderson’s geographically restricted calculations, the number of Broadway houses was relatively stable in the 1970s. A more capacious look at the industry might see the numerous new or converted peripheral houses as reshaping the definition of Broadway.

So were theater owners adjusting the supply of Broadway theaters in the 1970s? Maybe. And what market forces (if any) encouraged such changes? Unfortunately, I still do not know.

The Shubert Brothers/Shubert Theater Organization/Select Theaters Corporation/Shubert Foundation

Perhaps one problem here is that, looking back, we have a far stranger perspective on the industry than any theater owner might have in the moment. On the one hand, we have the benefit of hindsight and of seeing the entire field at once. On the other hand, we do not know the players, the petty jealousies, and the gossip that give histories like Riedel’s their punch. Maybe, to think like a Shubert, we have to see the world more as they did. Let us try then to narrow our perspective and look simply at their portfolio of theaters.

Today, the Shubert Organization owns 17 theaters on Broadway, each of which they have operated since at least 1965.14 We can see all the theaters’ occupancy over the past 50 years in Figure 7, with blue bars indicating runs of a musical, red for a play, and gold for anything else (solo show; comedy; magic).

Figure 7. Theater Usage at Shubert Houses, 1965–2014
The percentages next to the theater names indicate the total percent of days the theater was in use over the past half-century, though that number does not account for any periods when the theater was leased for non-legitimate stage use. Clearly, the large musical houses are the prime money-makers, with musicals running for ever-greater lengths (the Majestic’s rightmost bar is none other than The Phantom of the Opera). The mixed-use houses run from 68 to 78% full, though the Ambassador’s number is somewhat inflated by its current tenant, Chicago. And then the bottom four properties, small houses primarily for plays, only host a show about half the time.

We might look at this set of theaters and their overall occupancy on a seasonal basis to see better any trends over time (Figure 8).

Figure 8. Seasonal Occupancy at Shubert Houses, 1965–2014
This chart shows the percent of days each season every Shubert house (the light grey lines) was in use, with the red line giving the total occupancy percentage for all seventeen Shubert houses. From 1965 to 1975, the Shubert houses dropped from just under 70% average occupancy to just over 40%, with particularly bad years in 1969 and 1971 and 1972. The latter half of the 1970s was noticeably better, with almost all their theaters running above 50% in 1977, a remarkable feat. But after 1980 the next decade was one long decline, this despite the advent of mega-musicals such as Cats, Phantom, and Les Misérables, which all ran in Shubert houses. Michael Riedel, summarizing the view from the Shubert offices after the 1987–1988 season, emphasizes Broadway’s record-breaking grosses and 16% increase in attendance. Yet their overall occupancy rate was hovering just over 50% and was about to get significantly worse. For all the talk of a revitalized Times Square in the 1990s, the Shuberts were not running a “thriving theatrical empire” by the early 1980s.15

Or at least, the theater rentals aspect of that empire was not operating at peak capacity. Dark days were everywhere, particulary compared to the peak of 1980, when a mere 20% of the available rental days went unused by a show. But as Riedel observes, the Shuberts ran a complex, multi-layered business that could earn money even with unoccupied venues. For instance, in a more recent six-year period, the Shuberts earned $50 million from leasing their “air rights,” the right to the vertical space above their theaters.16 They also produced many of the shows in their theaters, which can be expensive and risky, but also guarantees that more returns go directly into Shubert pockets.17 On a smaller scale, the Shuberts earn income from selling concession rights. And of course, historically, they ran a near-total monopoly on theater booking and touring throughout the US for around 45 years, from around 1908 when the Syndicate died off until 1956, when the government won an anti-trust lawsuit that forced them to sell many venues and release their hold over road bookings.18

Moreover, the entire structure of the Shubert Organization meant that the flow of any income—even and including rental income—was always extremely complex and could disguise the size of the business’ actual income or loss. For instance, in 1924 the Shubert brothers issued $4,000,000 in bonds, followed by 150,000 shares of common stock. The announcement of the deal included a thorough accounting of the Shubert Theater Corporation’s recent revenues, as well as its assets and liabilities.19 Yet the offering failed to mention that most of the Shuberts’ real estate was not part of the corporation but rather leased to it by the brothers themselves.20 Therefore some substantial amount of the capital they sought (and successfully raised) to finance their theater operations and production business went directly back to the brothers as lessors of real property, even though it appeared as a cost to their corporation.

Things grew weirder after the brothers’ bankruptcy and their reorganization in the 1930s. Materials in the Shubert Archive from the 1940s, for instance, reveal leases and subleases, companies within companies: Shuberts all the way down. (Except for the 44th Street and the St. James Theatre, which sat on land owned by Vincent Astor, who was charging the Shubert’s subsidiary Select Operating Corporation rent equal to the annual real estate taxes plus 50% of net profits.21) By the 1970s, the Shubert Foundation, a non-profit founded after J. J. Shubert’s death, owned the Shubert theaters. In 1979, the IRS exempted the Foundation from normal rules that required charitable foundations to divest themselves of for-profit assets. So the Shubert Organization, as best I can tell, leases the right to operate Shubert Foundation-owned theaters, further obscuring the costs of bad business. Add in the complications of real estate taxes—of which we have all learned much in recent days—and it seems impossible to divine the Shubert Organization’s financial status at any given time. Precisely because the Shuberts have (or had) interests in so many different aspects of the theater business—including, at various times, producing, theater rentals, ticket sales, and tour booking—they can afford to lose money in some areas if those losses would enable greater gains elsewhere. In short, we should not expect everything (or even anything) about the Shuberts’ business to make sense financially because the financial incentives extend far beyond the supply and demand of theaters.

Dark Days

I set out to understand how theater owners manage dark days, the days in which their businesses operate at a loss. I hoped that a production’s grosses and the supply of theaters would help explain theater owners’ approach to dark days, particularly combined with a holistic view from a single executive suite. But I simply do not know enough to draw any conclusions. I have, however, laid out some of the difficulties in understanding the actual economic operation of the theatrical real estate. Among the factors discussed above are the divergent incentives for theaters and productions, the multiple forces affecting the supply of theaters, and the complexity of the corporations that own and operate theaters. Other variables not even considered in this paper include competition among theater owners (e.g., the Shuberts vs. the Nederlanders), differences in the types of theaters and their capacities (e.g., demand for musical houses vs. demand for smaller play houses), and the specific theater locations, even right around Times Square. All told, it seems far harder to exclude variables as irrelevant to owning theatrical real estate than it is to add more complications. I believe, however, that the project of understanding Broadway as in part a real estate business remains essential. Because theaters, those glorious, strange, edifices, remain among the scarcest resources in the profession, Broadway theater owners are some of the most powerful organizations in the American theater and have been for over a century. Their operations should not remain a black box.

1. Usually the stop clause amount must not be achieved for two consecutive weeks. Exemptions often apply for known bad periods. David M. Conte and Stephen Langley, Theatre Management: Producing and Managing the Performing Arts (EntertainmentPro, 2007), p. 99.

2. Plymouth Theatre, Real Estate Series, Shubert Archive.

3. Agreement between Jujamcyn and the Shubert Organization and McCann/Nugent, February 3, 1983, Shubert Organization Materials, T-Mss 2002-029, Box 1, New York Public Library for the Performing Arts.

4. Because a house’s gross capacity changes over time, for this chart I estimate the stop clause as 45% of the maximum gross from the previous year.

5. Conte and Langley, Theatre Management, p. 239.

6. The previous Simon play at the Plymouth, The Odd Couple, never even approached its stop clause when Simon transferred it to the Eugene O’Neill, which he co-owned. Mary C. Henderson, The City and the Theater: The History of New York Playhouses, A 250-Year Journey From Bowling Green to Times Square (Back Stage Books, 2004), p. 327. A memo regarding The Odd Couple’s booking at the Plymouth, dated July 28, 1964, notes that the profits for that show split in a manner more favorable to the Shuberts than those for Star-Spangled Girl. After The Odd Couple’s success, the next deal gave the Shuberts 40% less on the first $20000 of the gross (from 30% to 18%), though the split on the remaining gross remained the same, 25% for the Shuberts. Plymouth Theatre, Real Estate Series, Shubert Archive.

7. Ambassador Theatre Group, however, plans to open a renovated Hudson Theatre this season, bringing the theater count to 41. The Hudson has been out of the legitimate theater business since 1974.

8. In a more mathematically sophisticated version of this analysis, I hope to quadrangulate occupancy rates, the number of theaters in operation, production run lengths, and the time between occupancies.

9. Samuel L. Leiter, Ten Seasons: New York Theatre in the Seventies (Greenwood Press, 1986), p. 181.

10. Michael Riedel, Razzle Dazzle: The Battle for Broadway (Simon & Schuster, 2015), p. 102.

11. Leiter, Ten Seasons, p. 173.

12. Henderson, The City and the Theater, p. 197.

13. The Phoenix was located at Second Avenue and 12th Street, and thus sat far outside the Theater District. The Playhouse (later the Jack Lawrence), was both quite far west of most Broadway houses and was marginally below the 500-seat threshold that helps define Broadway theaters today.

14. The Music Box they co-owned with Irving Berlin’s estate until 2007., accessed October 2, 2016.

15. Riedel, Razzle Dazzle, pp. 369–360. One important problem with both of these charts: they underestimate the actual usage of each theater. Theater owners make the most money when shows are open and running, which is what this chart records. But shows pay some rent durng the load-in and technical period (usually around two weeks at least), load-out (another couple days or more), and, of course, previews (now seemingly endless). Those payments may be lower than payments during a run, but they are not actual dark days for the theater.

16. Ibid., p. 359.

17. The Shubert Organization returned to producing in the 1970s after a two-decade hiatus.

18. Peter A. Davis, “The Syndicate/Shubert War” in Inventing Times Square: Commerce and Culture at the Crossroads of the World (The Johns Hopkins University Press, 1991).

19. “The Dramatic Stage: Shuberts Float Bond Issue,” The Billboard 36, no. 27 (July 5, 1924), p. 20.

20. Jerry Stagg, A Half-Century of Show Business and the Fabulous Empire of The Brothers Shubert (Random House, 1968), p. 221.

21. Minutes of the Select Operating Corporation Board Meeting, 1941, Corporation Series, Box 60, Select Theatres Corporation/Select Operating Corporation Minutes, Shubert Archive.