How Effective Are Expert Tv Hosts at Saving Failing Businesses?

Home , Bar Rescue, Bar Rescue (season 1), Bar Rescue (season 2), Bar Rescue (season 3), Bar Rescue (season 4), Bar Rescue (season 5)

HOW EFFECTIVE ARE EXPERT TV HOSTS AT SAVING FAILING BUSINESSES?

RUSSELL S. SOBEL, REAGAN N. SOBEL, DOUGLAS M. WALKER and PETER T. CALCAGNO∗

The profit and loss system is an integral part of a dynamic market economy. Losses eliminate businesses that are inefficiently managed or whose products no longer provide sufficient value. Almost a dozen popular television shows feature entertaining expert hosts claiming to “save” failing businesses with injections of physical and human capital. We undertake the first comprehensive analysis of these shows, calculating the failure rates of the businesses and analyzing the incentive structure facing the shows, networks, hosts, and participants. In general, we find that these shows are largely unsuccessful in saving failing businesses. (JEL L26, L82)

[T]he choices TV programmers make about what gets differs in that the host actually purchases an made reflect more than their attempts to please audi- equity stake in the businesses featured. In the ences … TV is a business. That matters because the way TV conducts its business has a direct impact on others, the hosts spend a few days filming and the process by which programs are selected, financed, then move on to other establishments. The shows and produced. (Magder 2004, 138) also typically provide a healthy dose of entertaining family, employee, or owner infighting and I. INTRODUCTION drama, which are also aided by the host during their brief visit. The “reality TV” genre has become extremely In this paper, we provide the first comprehen- popular since Cops and MTV’s The Real World first aired in 1989 and 1992. To date, almost a sive analysis of these popular TV shows that por- dozen reality television (TV) shows have aired tray expert hosts claiming to save failing busi- featuring celebrities or experts attempting to nesses. Assessing the success of these shows is save struggling businesses. In shows such as no small task because of the necessity of the Kitchen Nightmares, Bar Rescue, and Tabatha’s large quantity of both qualitative and quantitative Salon Takeover, the hosts provide physical and data. To conduct our analysis, we compile data on human capital injections into the business, such every establishment that has appeared on all 11 as building renovations, training, new menus, different TV shows (over 500 total businesses), and rebranding. One of these shows, The Profit, examine thousands of actual customer reviews, conduct phone interviews with business owners, watch hundreds of hours of the actual shows, ∗The authors would like to thank Erin F. Carter for and compile first-hand accounts from owners, research assistance, and R.N.S. would like to recognize the financial support of the Center for Public Choice & Market employees, and local reporters. Process at the College of Charleston. We thank John Dove, At the outset it is important to note that Stewart Dompe, Michael Barth, Robert Lawson, Matthew many of the businesses featured on these reality Rutherford, Robert Salvino, Al Lovvorn, Thomas Duncan, Shelton Weeks, seminar participants at Florida Gulf Coast TV shows are failing because they probably University and the Southern Economic Association Meetings, should. They are depicted in the episodes as and three anonymous referees of this journal for helpful com- mismanaged, dirty, outdated, lacking value ments and suggestions. Sobel, R. S.: Baker School of Business, The Citadel, Charleston, SC 29409. Phone 843-953-5162, Fax 843-953 ABBREVIATIONS -6764, E-mail [email protected] Sobel, R. N.: Department of Economics, College of AAFR: Average Annual Failure Rate Charleston, Charleston, SC 29409. Phone 843-953-8100, AASR: Average Annual Survival Rate Fax 843-953-0745, E-mail [email protected] BLS: Bureau of Labor Statistics Walker: Department of Economics, College of Charleston, Charleston, SC 29409. Phone 843-953-4279, Fax 843- CCR: Cumulative Closure Rate 953-0745, E-mail [email protected] OLS: Ordinary Least Squares Calcagno: Department of Economics, College of Charleston, POS: Point of Sale Charleston, SC 29409. Phone 843-953-8192, Fax 843- TV: Television 953-0745, E-mail [email protected]

9 Contemporary Economic Policy (ISSN 1465-7287) Vol. 37, No. 1, January 2019, 9–24 doi:10.1111/coep.12285 Online Early publication March 30, 2018 © 2018 Western Economic Association International 10 CONTEMPORARY ECONOMIC POLICY propositions, offering poor customer service, auctioned, told some participants which lockers and/or targeting a dwindling customer base. We to bid on (and how much), and even financed believe there is good reason to be skeptical that a some cast members’ bids (Gardner 2013). The physical renovation and a few days of advice and house-hunting and renovation show Fixer Upper training—even by a talented expert—can alter was exposed for complete inaccuracy and out- the trajectory of a business’s performance. We right dishonesty by hiding the fact that the people expect the businesses featured on these shows to on the show were required to be under contract on have higher than average failure rates. a specific house, despite a large part of the show A vibrant economy will have both a large portraying them as house hunting and decid- number of new business start-ups and failures ing among several houses to potentially renovate (Sobel 2015; Sobel, Clark, and Lee 2007). Mini- (Foxnews.com 2016). The network responded mizing business failures should not be a society’s that viewers enjoy the show because of the inter- a priori goal—neither all businesses are worth action between the show’s hosts, not the accuracy saving nor can all be saved. In an economy where of the portrayals. entrepreneurs—even those with unusual ideas or Our analysis leads us to conclude that to a no prior skills—can open a business, there will large extent, because these “save the business” inevitably be many business failures. The profit shows are businesses themselves, the shows’ and loss system is an integral part of a dynamic hosts, producers, and networks are primarily market economy and it is characterized by churn- interested in making successful, profitable TV ing, referred to by Schumpeter (1934 [1911]) as shows and pleasing sponsors with placements. a process of creative destruction. Losses weed Accuracy and effectiveness are not the goals; out businesses that are managed inefficiently, or rather the shows are in the business of providing whose products do not provide sufficient value to entertainment to TV viewers to increase the view- justify production costs. ership exposed to the placements and advertise- Prior research has found the long-run out- ments of sponsors. Unfortunately, what makes comes of other reality TV shows intended to for a successful TV show may not be what is “help” people to be ineffective in a wide variety best for struggling businesses. Nor are the busi- of areas. For example, Fothergill et al. (2016) find nesses selected for the shows the ones most able that the reality weight-loss show Biggest Loser to be saved. This differential in incentives helps is not effective when subjects are examined long explain why the shows tend to have abysmal term. Similarly, the A&E show Hoarders has records at saving businesses. At best, due to the only mixed results in helping those with hoarding TV exposure, they mostly provide a temporary disorders when subjects are viewed over the long bump in revenue for an otherwise failing busi- term (Weiss 2010). Essentially, we are asking the ness. Consistent with our conclusion is the fact same question about these “save the business” that the one show where the host actually takes shows, and are the first to do so in a comprehen- an equity position in the business, The Profit, has sive and scientific manner. a much better record than the other shows as his The previous literature on reality TV shows incentives are the only ones truly aligned with the suggests that what is portrayed on TV is dis- long-run success and profitability of the business. torted for entertainment purposes; it cannot be taken at face value, and we confirm this bias in the shows we examine. For example, Oliver II. DEFINING SUCCESS AND FAILURE (1994), Doyle (2003), and Lawson and Lawson Chef Gordon Ramsay’s Kitchen Nightmares (2016) find that the reality police show Cops, was the first reality TV show to move into the one of the longest running reality TV series in domain of saving struggling businesses. Since history, distorts reality through highly selective then, a variety of other shows have aired. Table 1 editing and an unrepresentative sample of actual is a listing of the 11 American TV shows we crimes to boost viewership and advertising rev- analyze that attempt to save failing businesses. enue. For example, violent crimes and encoun- The shows are listed by original air date.1 ters with police are rare in actuality but are portrayed as widespread and common on the show because these scenes generate viewership. Simi- 1. Additional shows continue to be launched even while larly, the show Storage Wars faced litigation over this paper is under review, including Garage Rehab featuring Richard Rawlings saving failing garages on the Discovery rigging the auctions. Apparently, producers had Channel, and a new show from Tabatha Coffey helping family planted items in storage lockers before they were businesses on Bravo entitled Relative Success. SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 11

TABLE 1 American “Save the Failing Business” Television Shows, as of June 2016 No. of Estimated Seasons Viewers, in Show Title, Host’s Name Network Years Aired (First Air Date) (Episodes) millionsa Kitchen Nightmares, Gordon Fox 2007–2014 (September 19, 2007) 6 (92) 3.0 Ramsay Tabatha’s Salon Takeover / Bravo 2008–2013 (September 21, 2008) 5 (52) 0.9 Tabatha Takes Over,b Tabatha Coffey Restaurant: Impossible, Robert Food Network 2011–2016 (January 19, 2011) 13 (160) 1.7 Irvine Bar Rescue, Jon Taffer Spike 2011–present (July 17, 2011) 4 (113) 2.2 Hotel Impossible, Anthony Travel Channel 2012–present (April 9, 2012) 7 (89) 0.6 Melchiorri Hotel Hell, Gordon Ramsay Fox 2012–present (August 13, 2012) 3 (16) 4.4 (Fox) Bakery Boss/Buddy’s Bakery TLC 2013–2014 (May 27, 2013) 2 (13) 0.9 Rescue,c Buddy Valastro Save My Bakery, Kerry Vincent Food Networkd 2013–2014 (July 13, 2014 & March 9, 2014) 1 (10) 0.8 The Profit, Marcus Lemonis CNBC 2013–present (July 30, 2013) 3 (40) 0.6 Restaurant Redemption, Cooking Channel 2013–2015 (October 29, 2013) 2 (26) 0.1 Ching-He Huang Save Our Business, Peter Jones TNT 2014 (February 28, 2014) 1 (6) 0.7

Notes: Episodes include some re-visits to businesses saved on previous episodes, and other special episodes. CNBC, Cable National Broadcasting Company; TLC, The Learning Channel; TNT, Turner Network Television. aViewership data are averages of all available data sourced from TV by the Numbers (http://tvbythenumbers.zap2it.com), Son of the Bronx (http://sonofthebronx.blogspot.com), The Futon Critic (www.thefutoncritic.com), and Wikipedia (http://en .wikipedia.org). bTabatha’s Salon Takeover expanded to include more than just salons beginning in Season 4 and was renamed Tabatha Takes Over. cBakery Boss was renamed Buddy’s Bakery Rescue for the second season of the show. dSave My Bakery had a pilot episode that aired on the Cooking Channel on July 8, 2013; the remaining episodes were part of a full season that was broadcast on the Food Network starting March 9, 2014.

Four of these shows routinely captured over 1 TABLE 2 million viewers, and a few stayed on the air long Average Annual Failure Rates (AAFRs) by enough to have 75 or more episodes. However, Industry, 2000–2015 most were minor shows with only a handful of episodes and limited viewership. They have also All Firms 11.8% featured businesses in a wide variety of industries Leisure and Hospitality 10.7% including restaurants, hotels, bakeries, and even Food Service and Drinking Places 16.7% hair salons. We begin our analysis by discussing Accommodations 4.6% the shows, the annual business failure rates of the Source: Calculated from U.S. Department of Labor’s underlying industries, and the postshow failure Bureau of Labor Statistics (BLS), Business Employment rates of the establishments featured on the shows. Dynamics Database, compiled from longitudinal linking of Obviously, our analysis should account for Quarterly Census of Employment and Wages data (http:// the differing failure rates across these industries. www.bls.gov/bdm/charts.htm). Broad industry failure rates are available from the U.S. Bureau of Labor Statistics (BLS), Business their industries, there would be substantial Employment Dynamics Database. These failure differentials in the postintervention failure rates rates are provided in Table 2 for the industries for the different shows, simply based on the we examine. different industries in which they operate. For As noted in Table 2, the overall business example, Gordon Ramsay hosted two differ- average annual failure rate is 11.8%, while the ent shows in two different industries (Kitchen rate for accommodations is 4.6% and for food Nightmares and Hotel Hell) and our later service/drinking establishments it is 16.7%. results confirm the differential outcomes across Even if these expert hosts could bring these his two shows are entirely explained by this struggling establishments up to “average” for industry differential. 12 CONTEMPORARY ECONOMIC POLICY

It is important to note that the BLS data are for establishments from season i,orSi: √ annual failure rates. Some of the data we examine ( ) 1 for the shows are seasons that aired up to a ET ET Ni (2) S = Ni i = i . decade ago. This must be taken into account i EV EV in the analysis. If 10% of businesses fail each i i year, then in a sample of 100 randomly selected Because we have multiple seasons occurring businesses visited in a specific year 10 would at different times, with differing numbers of be expected to fail the first year, 9 more (10% establishments that we wish to weight equally, of the remaining 90) would be expected to fail an additional step is required. Letting Z be the in the second year, roughly 8 more in the third total number of seasons, we can weight each sea- year, and so forth. So if we witness only 73 out son’s survival rate by the share of establishments of 100 businesses surviving after 3 years have visited in that season relative[ to the∑ total num-] passed, the failure rate is not 27%, but rather Z ber visited across all seasons EVi∕ EVi to 10% annually. It is critical to be precise on these i=1 compute the weighted average. Thus, the com- two numbers and not use them interchangeably. bined, weighted, average annual survival rate We henceforth term the overall percentage closed (AASR) across all seasons is: as the cumulative closure rate (CCR), while the more accurate yearly figure we term the average ⎡⎛ ⎞ ⎤ annual failure rate (AAFR). ⎢⎜ ⎟ ⎥ ∑Z EV Popular press articles and TV segments about (3) AASR = ⎢⎜ i ⎟ · S ⎥ a few of these shows have provided grossly inac- ⎢⎜ ∑Z ⎟ i⎥ i=1 ⎢⎜ ⎟ ⎥ curate assessments of the outcomes because they ⎣⎝ EVi ⎠ ⎦ do not recognize this distinction. They typically i=1 cite the CCR, calling it a failure rate, without ⎡⎛ ⎞ ⎤ ( ) 1 making any adjustment for the time that has ∑Z ⎢⎜ ⎟ ⎥ EV ET Ni passed (Interrobang 2014; Satran 2015). Citing = ⎢⎜ i ⎟ · i ⎥ . ⎢⎜ ∑Z ⎟ EV ⎥ only the CCR is equivalent to saying the death i=1 ⎢⎜ ⎟ i ⎥ rate in the United States is 100% because every- ⎣⎝ EVi ⎠ ⎦ i=1 one eventually dies. It is a meaningless way to analyze the performance of the TV shows. The CCRs for two shows could be completely dif- The AAFR is simply 1 − AASR. This is our formula for calculating the failure rates of busi- ferent simply because of the time passed since 2 the interventions, even if the AAFRs were iden- nesses featured on these shows. These rates will tical. Presumably, press outlets and blogs rely on then be compared to the industries’ AAFRs. the CCR because computing the AAFR is more tedious. However, the AAFR is the more accu- III. ANALYSIS AND RESULTS rate representation of ongoing firm performance in the industry—and the appropriate one to use in In order to analyze the TV shows’ failure rates judging the relative performance of our different relative to those of the industry, we first compile “save the business” TV shows. a list of all business establishments featured on Since the shows occur over several years and the shows sourced from the TV shows’ websites seasons, we must properly weigh and compile the and checked with other sources, including TV data into an AAFR that accounts for the length Guide, other reality TV websites, and Wikipedia. of time that has passed since each season of the To determine if an establishment remains open show. Let EVi be the number of establishments we used data from Yelp, TripAdvisor, Google, visited in season i of the show, ETi be the number of establishments surviving today from season i, 2. As a numerical example, assume two seasons; season and Ni be the number of years ago season i aired. 1 being 5 years ago, 10 establishments visited of which 6 are still open; season 2 being 3 years ago, 14 visited and 8 still If Si is the annual survival rate for establishments from season i of the show, the relationship is open. The calculation would be: given by: ( ) 1 ( ) 1 10 6 5 14 8 3 AASR = · + · Ni . (10 + 14) 10 (10 + 14) 14 (1) ETi = EVi · Si = 0.860266 ≈ 86% Because we know the data for ETi and each EVi (and Ni) we can solve for the survival rate and thus, the AAFR = 1 − AASR = 1 − 86% = 14%. SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 13

TABLE 3 Kitchen Nightmares’ Failure Rate Details No. Usable Number Closed Establishment as of Percentage Season Air Dates Observations June 2016 Closed Season 1—Set 1 September 19, 2007–December 12, 2007 10 9 90.0% Season 1—Set 2 September 4, 2008–January 15, 2009 11 11 100.0% (Season 1 totals)a September 19, 2007–January 15, 2009) (21) (20) (95.2%) Season 2 January 29, 2010–May 21, 2010 11 9 81.8% Season 3 January 21, 2011–May 20, 2011 12 7 58.3% Season 4 September 23, 2011–March 30, 2012 14 9 64.3% Season 5 October 26, 2012–May 10, 2013 13 5 38.5% Season 6 April 11, 2014–September 12, 2014 6 4 66.7% Totals 77 54 70.1% Analysis of results AAFR for show 30.1% Industry-level AAFR (eating & drinking establishments) 16.7% *** Test statistic for Ho: show AAFR = industry AAFR 7.778 Ratio of show AAFR to industry AAFR 1.80

aThe two parts of Season 1 aired 9 months apart; some sources improperly count it as two separate seasons. Level of significance: * = .10, ** = .05, *** = .01.

Facebook, Instagram, the Yellow Pages, and the in 2014 after six seasons (Bricker 2014). The establishment’s website. In some cases, we called data for this show are given in Table 3. Overall, the business in question and interviewed own- 70.1% of the 77 establishments from the show ers.3 Our criterion needed to be clear and sim- have closed, and the inaccuracy of assessing the ple: Would a customer stopping by today find the show based on this CCR is clear in the table. establishment in operation, in that same location, As our earlier example illustrated, 100 businesses fundamentally as it was left by the show, even if with an annual failure rate of 10% implies a it reverted back to the original name, or if it was yearly path of 90, 81, 73, and so on, surviving sold to a new owner? businesses. It is easy to see the similar pattern in To ensure enough time had passed to assess the Kitchen Nightmares data. The failure rate is the possible business failures, for the few ongoing the highest for the first season, with only one of shows we ensured at least 1 year had passed from the 21 establishments remaining in business. the show’s visit for the results to be included. Using Equation (3), the AAFR for all Kitchen Once the data were compiled by season, we then Nightmares businesses is 30.1%, as shown in the aggregated them by year to use our formula from lower panel of Table 3. For context, Table 3 also Equation (3) to derive the AAFR for the show shows the industry failure rate of 16.7%. The across all episodes using the year of filming (not AAFR for the businesses featured on Kitchen air date) for our year coding.4 Nightmares is almost twice the industry average. Academically, the question becomes whether the difference between the industry average and the A. Example Results: Kitchen Nightmares show is statistically significant. We present the full details of our computations The null hypothesis for this test (Ho) is that for one show to serve as an example, but for the AAFR of the businesses featured on the the remainder we present only the results due show is equal to the industry’s AAFR. This to space constraints. Kitchen Nightmares ended distribution has mean E[Y] = p and variance var.(Y) = p·(1 − p)/n. Due to the small sample size, the critical values are calculated using the 3. As we will discuss, most owners sign legal agreements barring them from discussing their experience on the t-distribution with n − k degrees of freedom (see show. For a few of the shows, websites contain lists of Judge et al. 1988, 45).5 The resulting t-statistic the establishments and whether they are open, closed, or sold (www.foodnetworkgossip.com, www.barrescueupdates .com, www.grubstreet.com, www.theprofitupdates.com, or 5. The number of observations used to compute the www.realitytvrevisited.com). However, these are not always degrees of freedom is not simply the number of establish- updated so we checked each establishment independently. ments. Because these are annual failure rates, and the business must survive each year’s cut, the true number of observations 4. Some establishments spanned two episodes. To avoid is instead the number of establishment-years (i.e., the sum of double counting we assigned these to the first episode on the number of establishments for each season times the num- which they appeared. ber of years over which it is observed). 14 CONTEMPORARY ECONOMIC POLICY

(7.778) with the indicated level of statistical viewed as an alternative measure by which the significance is also shown in Table 3. Indeed, success of the different shows could be com- the average annual failure rate of the businesses pared. The problem with this, however, is that the featured on Kitchen Nightmares is significantly postintervention reviews may suffer from a bias. higher than the industry, at a 1% level. Thus, This is because some customers of a restaurant while the record of the show at “saving” failing “saved” by Gordon Ramsay, for example, may businesses is not nearly as bad as portrayed in have higher expectations than they would have the popular press, which often cites the CCR of prior to the intervention, which could bias the 70.1%, the AAFR of show participants is still postreviews. The prereviews are not subject to significantly higher than the industry average. this bias. Thus, our primary use is of the prere- Kitchen Nightmares’ AAFR of 30.1% divided view data as a measure of customer satisfaction, by the industry average failure rate of 16.7% which is known from prior literature to be closely yields a ratio of 1.80, also shown in Table 3. A correlated with business’ success rates. ratio of 1.80 means the AAFR from the show is These data were difficult to obtain because we 80% higher than the industry average. In other needed to pinpoint the exact date of the actual words, even after being on the show the estab- visit and intervention (not the date the show lishments are 1.8 times as likely to fail as another aired), and we needed a sufficient number of pre- randomly selected business in their industry. It is reviews and postreviews. It required coding every worth explicitly noting that had the show’s fail- customer review (pre or post show visit) and aver- ure rate been lower than the industry average, aging them separately. We did this for a subsam- the ratio would be less than one. Since this ratio ple of 125 establishments over 2013 and 2014, accounts for different industry failure rates, we the two common years all 11 shows aired. This use this ratio to compare the effectiveness of dif- involved hand coding almost 12,000 individual ferent shows in different industries. customer reviews (4,366 pre; 7,401 post). The number of reviews differs by establishment, aver- aging 35 preintervention and 59 postintervention B. Customer Ratings reviews. One concern in comparing the postintervention failure rates across shows is whether C. Results across Industries the shows were intervening in businesses that We present our main results in Table 4. The were equally likely to fail without help. It is likely shows in the table are sorted by industry, and that any business agreeing to appear on a “save ranked within industry to aid in comparison. We the business” show is more likely than average organize the following discussion by industry. to fail. Still, this could differ across shows, especially The Profit who is attempting to select good Food Service and Drinking Establishments. The businesses in which to invest. To gain insight we top section of Table 4 contains the results for the collected a sampling of preintervention customer shows that intervene in food service and drinking reviews (“prereviews”) for the establishments establishments. Restaurant Redemption, featur- from Yelp and TripAdvisor, where any customer ing chef Ching-He Huang helping failing Asian can rate an establishment on a scale of one to five restaurants, has the fewest usable observations at stars, with five being the highest. A strong cor- 26, but each of the others has 77 or more. There is relation exists between business success/failure a striking equality in the results from Restaurant and customer ratings (Chevalier and Mayzlin Redemption, Kitchen Nightmares, and Restau- 2006); the lower the prereviews, the higher the rant: Impossible, all with AAFRs of roughly probability of failure without help. While there 30%, and all with failure rates significantly higher is no perfect way to judge how close a business than the 16.7% industry average, at the 1% level. is to shutting down, the prereview averages for The ratios of the show to industry failure rates the establishments are the best proxies available are 1.77, 1.80, and 1.80, respectively, implying to reflect these differences for businesses across AAFRs for the establishments on the shows that the different shows. are roughly 80% higher than the industry average, We are also able to compute average post- even after the intervention. In essence, all three intervention review scores (“postreviews”) to shows are equally (in)effective at saving these analyze any change in review scores after the establishments. There is no differential across the show interventions. An improvement in customer three hosts, despite the differences in their back- review averages after the interventions could be grounds and interventions. SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 15

TABLE 4 Show Failure Rates Show Title Average Annual Failure Rates (AAFRs) Customer Review Scoresa Sample Size Show Industry Difference Ratio Sample Prereview Change (firms) (S, %) (I, %) (S − I, %) (S/I) Size (firms) Scores in Average Food & Drinking Establishments Bar Rescue 97 16.6 16.7 −0.1 0.99 20 3.18 −0.24 Restaurant Redemption 26 29.6 16.7 +12.9*** 1.77 18 3.09 0.17 Kitchen Nightmares 77 30.1 16.7 +13.4*** 1.80 20 3.45 −0.11 Restaurant: Impossible 133 30.1 16.7 +13.4*** 1.80 20 3.47 −0.11 Bakeries Save My Bakery 10 3.5 16.7 −13.2 0.21 5 4.44 −0.61 Bakery Boss/Buddy’s Bakery 13 19.2 16.7 +2.5 1.15 9 3.83 −0.31 Rescue Hotels Hotel Impossible 80 3.6 4.6 −1.0 0.78 20 3.29 0.43 Hotel Hell 12 8.5 4.6 +3.9 1.85 10 3.56 0.18 General/Miscellaneous The Profit—Deals 21 0.0 11.8 −11.8** 0.00 5 4.23 −0.86 The Profit—No Deals 9 0.0 11.8 −11.8 0.00 4 3.66 0.13 Save Our Business 6 5.9 11.8 −5.9 0.50 4 3.54 −0.04 Tabatha’s Salon 50 8.5 11.8? −3.3 0.72 14 3.78 −0.15 Takeover/Tabatha Takes Over

aThe number of customer reviews differs by establishments; see text for details. Customer review scores can range from 1 (lowest) to 5 (highest). Level of significance: * = .10, ** = .05, *** = .01.

In this group, Bar Rescue appears to have the unable to independently confirm these numbers. best record, with a show AAFR of 16.6%, vir- However, given the number of establishments in tually identical to (and not statistically different NAICS code 722410, drinking places, this would from) the industry average, resulting in a ratio imply an AAFR of over 11.5% based on “over of 0.99. The “prereviews” in Table 4 raise doubt 5,000” bars, and 14.9% based on “6,500.” The that higher initial quality among the establish- 16.6% AAFR for Bar Rescue is not significantly ments explains Bar Rescue’s lower failure rate. different from 14.9%, but is significantly higher This is because Bar Rescue’s establishments have than 11.5%, at the 1% level. Thus, Bar Rescue’s the second lowest prereview scores among these own industry data suggest that bars may have a four shows, at 3.18 (out of five possible stars). lower failure rate than restaurants, partly account- Interestingly, the virtually identical failure rates ing for the lower failure rate of this show. for Kitchen Nightmares and Restaurant: Impossi- ble is equally matched by their virtually identical Bakeries. Next, we examine the two shows that prereviews of 3.45 and 3.47, respectively. Per- focus on saving failing bakeries. The sample sizes haps most striking is that for three of the four are small at 10 and 13 establishments, and we shows, including Bar Rescue, the average cus- are reluctant to draw strong conclusions based on tomer rating falls after the intervention. the small sample sizes. Again, since we do not While Bar Rescue seems to clearly have the know the failure rate for bakeries, we have no best record among this first group, we must rec- real comparison group other than the food and ognize it is not really in the same narrow indus- drinking industry. The 10 bakeries featured on try as the other three. The BLS data only give Save My Bakery do have a lower failure rate than the failure rate for eating and drinking places the food and drinking industry at large, but with combined, and bars may have a different failure the extremely small sample size this differential rate than restaurants. At the beginning of some is not statistically significant. Bar Rescue episodes the narrator makes a state- An examination of the prereviews suggests ment about how many bars fail annually. How- that the lower than industry failure rate for estab- ever, these numbers have changed through the lishments featured on Save My Bakery may sim- seasons, from “over 5,000” to “6,500.” We were ply be a function of the bakeries being in better 16 CONTEMPORARY ECONOMIC POLICY shape prior to the intervention than other fail- with so few observations the difference from the ing businesses appearing on these shows. The industry failure rate is not statistically significant. average prereview is 4.44, out of 5, which is Tabatha’s Salon Takeover featured Tabatha surprisingly high. Similar to the first group of Coffey, an Australian hairstylist and salon owner, shows, these interventions also result in a decline helping failing hair salons and barbershops. Dur- in the average customer review scores, and the ing seasons 4 and 5, the show expanded to a wide −0.61 decline for Save My Bakery is relatively variety of businesses, ranging from restaurants large. It would seem odd that a show with signifi- and bars to a dog daycare facility. The show’s cantly better results would have one of the largest name was changed to Tabatha Takes Over.Again, reductions in customer reviews, so we believe the the BLS data do not have a failure rate specifi- evidence suggests that Save My Bakery’s better cally for salons, and since a wider variety of busi- record is explained by the businesses not being nesses were helped in later seasons, we compare initially as close to failure (or it could be evidence the results of this show to the overall AAFR for of bias in the postvisit reviews). However, again all industries. There is no significant difference. with such small samples, we are reluctant to draw Our most interesting results come from The firm conclusions. Profit. This show is different because host Mar- cus Lemonis personally invests in the businesses Hotels (Accommodations). Hotel Impossible’s profiled on the show. These include a variety of failure rate is slightly below the industry aver- establishments, selling such things as barbecues, age, while Hotel Hell’s rate is almost twice candles, shoes, jeans, signs, and furniture. There- the industry average, but neither difference is fore, we benchmark his results against the overall statistically significant. The lack of statistical AAFR across all industries. significance for Hotel Hell’s higher failure rate It is important to stress that, unlike the other shows, not all the businesses featured on The is because of the small sample size of only 12 Profit are actually failing, despite the show’s observations. Interestingly, the hotel industry claim that it is helping “struggling” businesses. (both shows) is the only one in which customer This is the only show for which we can precisely reviews uniformly increase after intervention. document a business’s financial stress. As part Interestingly, Gordon Ramsay is the only host of each episode, the host goes over the financial with shows in two industries. The two shows reports for the business, so in 33 of the cases it (Kitchen Nightmares and Hotel Hell)havediffer- was possible for us to determine the exact profit ent failure rates, 30.1% and 8.5%, which raises and loss levels of the businesses. Of these 33 for doubt that any particular host would be equally which we could identify the data, 18 (54.5%) successful working across industries. In fact, the were losing money, 2 (6.1%) were roughly differential in failure rates seems to be entirely breaking even, and 13 (39.4%) were profitable, 5 explained by the industries. Table 4 indicates that of which had annual profits exceeding $250,000 Ramsay’s shows have very similar ratios (1.80 per year. The profitable ones were often char- and 1.85). Thus, relative to the industries, Gor- acterized on the show as needing help because don Ramsay is equally (in)effective at saving fail- they had too much debt, or had other issues in ing businesses, having AAFRs about 80% to 85% production, distribution, or sales. higher than the relevant industry averages. Given the host Lemonis invests his own money, and he provides an ongoing postshow Other Industries. The final category includes interaction with the business owner, this show shows that claim to save failing business in a creates different incentives than the other shows variety of different industries. Save Our Business in our study. Since the host takes an ownership with host Peter Jones, a British entrepreneur and stake in the business, his incentives for the busi- businessman, featured a wide range of estab- ness’s survival and profitability align with the lishments across the United States, including an original business owners’. Therefore, we hypoth- indoor playground, karate school, floral shop, esize the failure rate of the businesses featured on furniture and mattress store, and bagel deli. Given The Profit should be the lowest among the shows. the wide variety of businesses, we compare the Interestingly, this show also creates a control AAFR of this show to that of all U.S. businesses, group, as Lemonis does not reach an agreement to 11.8%. The show aired for one season with six invest in all of the businesses shown. For this rea- episodes, and only one of those establishments son, we separate the data for The Profit into a pair has failed, resulting in a low AAFR. However, of results, one for the establishments he makes a SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 17 deal with and one for those he does not (indicated The prereviews are highly significant at the “Deals” and “No Deals” in Table 4). Out of the 30 1% level, indicating the higher the customer rat- businesses in our sample, 9 never reached a deal, ings prior to the intervention, the higher are even after the show aired. However, in some cases the postreview ratings. The coefficient suggests he does help the business prior to the deal falling that for each point higher is the prereviews, the through (The Profit Success 2015). postreviews are 0.519 greater. None of the year The results from The Profit are indeed differ- dummies is statistically significant. Of the show ent, as the AAFR is 0% for the businesses with dummies, only the two hotel shows, Hotel Impos- which he does make a deal. This rate is lower sible and Hotel Hell are statistically significant at than the overall U.S. average at a 5% level of the 1% and 5% levels, respectively. The coeffi- statistical significance, and The Profit is the only cient estimates suggest that businesses featured show with such a result. Interestingly, the failure on these two shows have postreviews that are rate of the “No Deals” businesses is also 0%. 0.864 and 0.764 higher, respectively. Recall that This may suggest that the host chose businesses the hotel industry data from Table 4 indicated that for the show that he expected to be successful, the average review scores rose only in the hotel regardless of whether they ultimately entered industry. This, coupled with the fact that reviews into a deal with Lemonis.6 Table 4 also shows scores did not increase for Ramsay’s restaurant the prereviews for businesses on The Profit are show, leads us to believe that this is an industry- higher than the businesses featured in most of specific rather than a host-/show-specific result. the other shows, consistent with these businesses To further test the effect of the shows on these being in better shape prior to the show. businesses, we test a Probit model of whether the business was still open as of the time our data D. Regression Models were collected in 2016. This variable is called Open, and coded witha1ifthebusinesswasstill Overall, our results thus far suggest that reality open and 0 if not. Again, because most of the vari- TV shows generally tend to be unsuccessful at ables are binary, this model contains significant attaining postintervention failure rates even as multicollinearity, making it difficult to include good as the industry average. In fact, for the all of the shows in this model. Independent vari- shows with the largest number of observations, ables are included for the four shows for which the results seem to suggest the establishments we have the most data: Bar Rescue, Kitchen featured on these shows fail at a rate almost twice Nightmares, Restaurant Impossible, and Hotel as high as the industry average. We now subject Impossible. We include show dummies, prere- our data to regression analysis. view scores, industry-level AAFRs, and dummy Because of the close correlation between variables for 2010–2013 to control for economic customer reviews and business success found conditions over time. Finally, we included a vari- in prior literature, we first construct an ordinary able (Time Passed) that counts the number of least squares (OLS) model of the postreviews months between when the show episode aired for the businesses. To what degree are the estab- and when we collected the data. The calculated lishments’ postintervention reviews explained by marginal effect coefficient results of this model the preintervention reviews versus the different are shown in Table 5, column B. TV hosts’ interventions? As noted above, we Only two variables are statistically significant calculated the average postreviews using Yelp, at explaining the probability of an establishment Facebook, Google, and Trip Advisor, for as remaining open. The coefficient on prereviews many of the businesses as we could. We then (positive and significant at the 1% level) indicates regressed these scores on dummies for individual that the higher rated the establishment was prior shows, the prereviews, and year dummies. The to the intervention, the more likely it is that the results from this regression are shown in Table 5, establishment is still open as of our data collec- column A. tion. For each one point higher was the prereview rating, the business is 16 percentage points 6. However, this does not mean there are not differences between the two groups in important areas we cannot assess, more likely to remain in business. In addition, such as revenue or sales growth. Since our original analysis, the industry AAFR is negative and statistically Amazing Grapes, a business on The Profit that entered into significant at the 10% level, suggesting that the an agreement with Lemonis, closed in July 2017, reportedly after problems with lawsuits and differences with Lemonis. higher the industry’s AAFR, the less likely the However, updating this result only slightly changes the overall establishment is still operating. The effect is quite show’s AAFR, and does not affect our conclusions. large as for each 1 percentage point higher is the 18 CONTEMPORARY ECONOMIC POLICY

TABLE 5 OLS Model of Postrating and Probit Model of Open Independent Variables OLS Model Probit Model (Marginal Effects) Dep. Variable: Postreviews Dep. Variable: Open Constant 1.174 (0.438) Show dummies Bar Rescue 0.221 (0.332) −0.034 (0.137) Hotel Impossible 0.864*** (0.325) 0.039 (0.230) Kitchen Nightmares 0.505 (0.332) −0.002 (0.143) Restaurant Impossible 0.407 (0.320) −0.109 (0.129) Hotel Hell 0.764** (0.344) — Save Our Business 0.489 (0.416) — Tabatha Takes Over 0.569 (0.370) — Restaurant Redemption 0.487 (0.326) — Save My Business 0.348 (0.389) — Buddy’s Bakery 0.369 (0.345) — Prereviews 0.519*** (0.081) 0.163*** (0.062) Industry-level AAFR — −0.031* (0.017) Time passed — 0.003 (0.008) Year dummies 2009 0.260 (0.661) — 2010 0.122 (0.434) −0.372 (0.464) 2011 −0.328 (0.249) −0.373 (0.360) 2012 −0.247 (0.215) 0.034 (0.244) 2013 −0.036 (0.132) −0.030 (0.134) N 145 145 F 4.53 — R2 0.3613 — LR—χ2 — 28.64 Pseudo R2 — 0.1544

Note: Standard errors are shown in parentheses. Level of significance: * = .10, ** = .05, *** = .01. industry failure rate, the business is 3 percentage surprisingly ineffective at actually saving busi- points less likely to be in business. Both of these nesses. Why? To a large extent the answer is are intuitive results. None of the show dummies the same as for the ineffectiveness of other real- has predictive power in the Probit model.7 ity shows on weight loss and hoarding. While Overall, the results of these two models bolster the shows explicitly claim their purpose is sav- the conclusion that the expert hosts are fairly inef- ing failing businesses, in actuality their primary fective at “saving” businesses. The primary deter- purpose, as for all other TV shows, is to provide minants of whether an establishment featured on profit for the production company and TV net- these shows remains open are: (1) the failure rate work that airs it (Blumenthal and Goodenough for the industry, and (2) the preshow quality of 2006; Mogel 2004). Saving failing businesses the businesses as reflected in their preinterven- is not the primary purpose of these reality TV tion customer reviews. We find no systematic shows. The shows themselves are business ven- or robust evidence that the shows themselves tures by production companies and TV networks. matter, or that the shows produce results that Whether a particular show is profitable depends differ in any way attributable to the host. on viewership data, which are closely linked to the ability of the shows to sell paid advertise- ments and sponsorship placements in the pro- IV. COMMON ELEMENTS OF “SAVE THE grams. Many of the save the business shows share BUSINESS” TV SHOWS common ingredients and problems, and we dis- Despite the longevity and viewership of some cuss some of these in this section. “save the business” shows, most of the shows are A. Differing Incentives

7. We did attempt to include other show variables and also When decisions are made of which businesses modeled the shows individually. Regardless of the specifica- to select for the shows, what to change/improve, tion, none of the show variables was significant. how to improve it, and on which problems to SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 19 focus attention, the incentives of the shows do FIGURE 1 not align with truly saving failing businesses. Shows’ Failure Rates versus Viewership Because the decision authority on changes rests with the show, and the upfront legal contracts 5.0 developed by the show are strong and binding, 4.5 when these incentives do not align it is the incen- 4.0 3.5 tives of the show that dominate. 3.0 It is difficult to speculate on exactly how 2.5 the business owners’ interests may vary from 2.0 the shows’ hosts and producers. This is because 1.5 1.0 the businesses featured on the shows are often 0.5 required to sign lengthy legal contracts that pre- (in millions) Show Viewership 0.0 vent owners and employees from discussing the 0.00 0.50 1.00 1.50 2.00 Failure Rate Ratio of Show to Industry show and what was performed.8 There are well- documented cases of owners being threatened with legal action if they talk about their experience on, or disparage, the show.9 Some may face imminent failure and simply want a tempo- maximize the long-run profitability of the busi- rary bump in revenue to postpone the inevitable, ness are strengthened by the differing results for while others may want to try to find and secure The Profit. a buyer for the establishment. In our interviews, In general, to align the incentives of the busi- some even claimed the show approached them to ness owners and the other shows, there would appear in an episode. need to be a link between viewership and the Again, The Profit is sufficiently different in survival rate of the businesses featured on the that the host personally invests his own money show. If there were a clear link, hosts would have in the business by taking an equity share, and a strong incentive to save the businesses, as it provides long-term managerial support. This cre- would directly increase viewership, ratings, spon- ates an alignment of a personal incentive for the sorships, and revenue for the show itself. show host that is linked to the future survival of Unfortunately (for some of the business own- the businesses, as well as an incentive to select ers), the opposite appears to be true. Figure 1 higher quality establishments to be featured. The presents the relationship between the ratio mea- AAFR of firms featured in The Profit is much sures for the show failure rates (from Table 4) lower than the industry average (and the other and the average show viewership (in millions). shows we studied). Our conclusions about the The relationship is clearly positive with a correla- importance of the different incentives between tion coefficient of 0.56. A simple OLS regression what is performed to create an entertaining and also results in a statistically significant relation- profitable TV show versus what would actually ship. That is, the “save the failing business” TV shows with higher business failure rates, and thus 8. For example, The Profit casting (http://www worst records, have the highest viewership. Thus, .theprofitcasting.com/) contains conditions such as “I viewership does not appear to increase from hosts shall keep in strictest confidence and shall not disclose to any being more effective at saving businesses, in fact other applicant, participant or other third party at any time it is the opposite. This is hardly surprising, given (i.e., prior to, during, or after the taping or exhibition of any episode of the Program) any information or materials of any the more drama and shocking behavior present kind, including without limitation...” the more entertaining is the episode, which likely 9. Ryland (2013) documents the owners of one establish- leads to shows selecting businesses with these ment threatened with a letter stating “We understand that you traits and major issues that simply cannot be are planning a public event … at which you will discuss your experiences and your ‘unflattering portrayals’ on the solved with a quick TV remodel. show. If you speak about the show without Upper Ground’s One major example of the differential incen- and Fox’s prior approval, and if you disparage the show, its tive is that some of the problems the hosts “solve” host, or its producers, you will breach your obligations under Paragraph 10 of your Personal Release and Paragraph 14 of during the TV episode are not really the prob- your Participant Agreement ...[t]hese agreements prohibit you lems hurting the business. Similarly, the show from speaking publicly about Kitchen Nightmares, other than hosts sometimes ignore key problems with the to acknowledge ‘the mere fact of your participation in the Series in personal publicity relating to yourself.’ Your con- businesses, while focusing instead on contrived duct exposes each of you to liability for liquidated damages problems that are created when show hosts pack of $100,000.” the establishments to capacity with customers as 20 CONTEMPORARY ECONOMIC POLICY a “test” for issues to generate entertaining TV Do the businesses really need new POS sys- footage of stressed employees. tems? Does a new menu of cocktails featuring In many first-hand owner accounts and user one particular brand of liquor help turn around reviews on Yelp, a frequent complaint is that a failing bar? Are these the best uses of the lim- the show does not renovate the bathrooms even ited resources at the show’s disposal to help the though they were in dire need of repair (Bar business? Perhaps not; the brands appear on the Rescue Updates 2014). Restaurants and night- shows because they are paid sponsors that must clubs are turned into five-star environments in be featured, and those changes are likely made the main areas, without a single change made to regardless of need; while other real needs may go nonfunctional, broken, or utterly disgusting cus- unsatisfied and unfixed. tomer restrooms.10 In another case, a pot-hole laden parking lot severely damaging customers’ C. Financial/Physical and Human Capital cars was not even mentioned or fixed in the show, Despite our findings, there is no question that despite online customer reviews both before and many of the businesses featured on these shows after the show citing it as the main reason the 11 have problems, and the shows do indeed provide business is not attracting customers. When lim- some help.12 The monetary investments made ited show resources are deployed, what the busi- into the business can certainly help to improve ness actually needs is of secondary importance their equipment, furnishings, building, help pay relative to the incentives to produce a profitable off debt, and expand physical capital (and/or TV show. Shows logically focus their scarce bud- make the business more attractive to a poten- gets on the things that generate high levels of tial buyer). The human capital invested by the sponsorship placements visible on TV, which are hosts can include training the business owner brand name logos and sponsor placements in and employees, changes in staffing, and changing areas like the dining area, kitchen, or hotel lobby. products and pricing. If these adjustments remain in place after the B. Promotional Placements show, there is a potential to change the outcome for the business. Simply changing the culture of The power of paid sponsorships is strong; employees within a business can have a dramatic it drives much of what is portrayed and which impact (Katzenbach, Steffen, and Kronley 2012; interventions are employed on the shows. For Schneider 1987). A few shows even engage the instance, the shows remove or cover existing employees and owner in teamwork and leader- beer taps, signs, or posters that show nonsponsor ship activities. Whether this human capital can brand names so their logos will not appear on the be communicated and permanently internalized screen in competition with the brands of the spon- by the owners and employees, given the limited sors (Tran 2012). Bar Rescue’s Jon Taffer updates time spent by the hosts (ranging from 2 to 7 days), drink menus with sponsor-branded liquors and however, is unclear, especially given the first- beers, for example; trains staff with employees hand accounts documenting that some hosts spent from sponsoring companies or distilleries; and no time with the owners or employees outside of installs high-tech tap and point of sale (POS) sys- what was on camera.13 tems of sponsors. Tabatha Coffey of Tabatha’s The ability to save failing businesses, in the- Salon Takeover/Tabatha Takes Over and Buddy ory, would be a skill reflected in the human capital Valastro of Bakery Boss/Buddy’s Bakery Rescue of the host. As such, one might expect this human also frequently install branded POS systems. capital to be improved through time and experience. However, even with empirical analysis we 10. A referee pointed out that this is missing an opportu- can find no evidence that the hosts became more nity for a new sponsor placement for restroom facilities. We successful in later seasons of their shows in our agree, but note that these shows may not be the best choice data. From season to season they repeat the for- for their advertising budgets, relative to other avenues. mula of picking disgusting or outlandish places 11. See Bar Rescue Updates (2015), Second Base Bar and Grill Update, which discusses a gang shooting at the nightclub that was not mentioned in the TV show; and Taraborelli 12. See, for example, Lore (2013) and Offitzer (2013), (2011) for a discussion of the cross-dressing hostess and although note the business discussed in Offitzer (2013) has lesbian, gay, bisexual, transgender customer base that was closed despite the glowing praise about the show’s “miracle.” driving away customers omitted from discussion on Kitchen 13. See, for example, Morabito (2010) for an owner Nightmares’ DownCity episode. Also, see online reviews of account, and Phu (2013) for a quote from one show host Stein Haus Brau & Brats (Friar Tuck’s) regarding the parking saying, “I isolate myself. I do not speak to anyone unless the lot. cameras are rolling...” SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 21 with interesting employees and owners, creating short days, and plan and put into effect a new problems to solve, appearing to solve the prob- business model in an unfamiliar town. The fact lems, and moving on to the next episode. that the majority of businesses actually experience worse customer reviews after the inter- 15 D. Publicity Effects and Local Tastes ventions is consistent with this problem. The hosts sometimes claim in interviews or follow-up Interviews with owners of the surviving busi- episodes that the failures after show are caused nesses suggest that the one helpful thing that the by owners not following through on the changes, shows did provide was publicity from the TV of which many do not. To temporarily avoid this, show airing. Even if the show simply prolongs the Kitchen Nightmares contractually obligates the time it would have taken for the establishment to establishments to serve the new menu for a few close, that may be a benefit we cannot measure months after the show, in an attempt to prevent precisely in our failure data. Thus, to the extent the reversion (Walthall 2011). Our analysis pro- that these shows do actually help it may be only vides some insight on two reasons why establish- due to a publicity effect from the establishment ments sometimes abandon the changes made by appearing on TV.14 Tomicki (2011) lists data for the shows. The owners and employees in some restaurants on several different TV shows, and cases feel like the changes were forced on them, argues the effect is roughly a doubling of business so they do not take ownership of the new ideas; in the short run, while Platt (2015) finds a slightly and because the changes result in complaints and smaller effect that is greater in the short term lost business from existing regular customers, than long term. Not all benefitted from this expo- hurting their postshow revenues. In any case, it sure, however, as some businesses that were por- would make sense to abandon changes which trayed negatively on the shows (in terms of rude have clearly led to declining sales or to alienating employees and owners or cleanliness problems) a large segment of regular customers. experienced substantial declines in revenue. In some cases, business owners have been quite bit- E. Contrived Personal Drama ter about their appearance on the shows and say The dynamics that make the best TV drama for they wish they had never participated. viewers include argument and infighting among The publicity of appearing on one of these employees; owners and employees resistant to shows may come with other negative effects. change; and incompetent owners that do not care Changing too many things at once, too quickly, about customers or cleanliness. It seems that the and having these changes selected and imposed shows pick failing businesses, and let personal by an outsider with different motivations can drama play out in a way that contains elements of result in disenfranchisement and a lack of own- both surprise and suspense. These are important ership in the ideas and discontinuance of the elements of entertainment demand (Bizzozero, changes implemented during the show’s interven- Flepp, and Franck 2016). tion. Major changes can also result in lost busi- One example of this is that in several of ness from loyal, long-time customers. This was the shows, the host visits and tries the products the main conclusion of Segal’s (2012) New York (food, cocktails) prior to implementing changes. Times article that attempted to follow up on a In almost every episode of Bar Rescue, Restau- handful of the Restaurant: Impossible interven- rant: Impossible, and Kitchen Nightmares,an tions. It is also frequently mentioned in owner early segment of the show focuses on the host (or accounts and on-line reviews (Klein 2014; Mora- his representatives) trying out food or drink selec- bito 2010; Taraborelli 2011). Postintervention tions from the menu. Almost without exception, customer reviews often complain about favorite the host comments on how terrible the food/drink menu items no longer being available, or a new is, in some cases spitting out or vomiting food. ambience different from what they like or were The episode typically ends with customers com- used to prior to the show’s visit. menting on how much the food or drinks have One frequent complaint regards hosts imple- improved as a result of the intervention. menting changes that are not consistent with local tastes (Dunn 2014; McGee 2013). It is obviously 15. The fact that the hotels do not suffer the review difficult for an outsider to come in, spend afew decline is also consistent with this. Hotels do not cater to local tastes or preferences, but to those of diverse travelers. It is easier for a well-traveled host to know what travelers prefer 14. In addition, according to Cieslak (2016) at least one than to know what the residents of a specific town would show actually pays the businesses for appearing on the show. prefer to eat. 22 CONTEMPORARY ECONOMIC POLICY

In other cases, people are sent into establish- relationship drama that can subsequently be ments to order cocktails that the bartenders do “solved” by the show’s host. not know in an attempt to embarrass employees. These encounters are used as examples of V. SUMMARY poor training, and the host then focuses on fixing these problems that are created, rather than the In this section, we have highlighted some ones actually plaguing the business. Saving a fail- of the elements common to “save the business” ing business requires assessing the true problems TV shows. Many of the shows follow a com- and allocating the scarce resources available to mon format, which includes the selection of repair and fix the most important issues. Having interesting and dysfunctional businesses often a successful TV show requires creating dramatic with tumultuous personal relationships among scenes and allocating resources to highlight the owners and/or employees, subsequently flooding products of the show’s sponsors. the businesses with more customers than they Several of the shows also feature “stress tests” can reasonably handle as a “stress test,” and prior to the renovation, in which the business is highlighting selective fixes that feature product flooded with substantially more customers than placements that often are not actually solving is normal. As would be expected, the establish- the underlying problems being experienced by ment often runs out of plates and glassware, the business. has abnormally long ticket/service times, quality Our analysis suggests that businesses partic- and cleanliness standards decline, and employees ipating in these shows have failure rates about (and owners) become stressed and overwhelmed. twice as high as their industry averages. We do The situation often creates dramatic TV scenes of not think the shows cause the higher failure rates. customers complaining and employees and own- Rather, preintervention customer reviews suggest ers arguing. These are then portrayed as the prob- that the businesses were more likely than average lems needing to be fixed by the host, and hours of to fail, and the TV show hosts did little to change training are then conducted as part of the episode this outcome. From the show’s perspective the to involve the employees to serve large num- dire situations faced by these businesses make for bers of customers more rapidly. The issue is that entertaining TV, and drawing viewership is the these are fundamentally not the normal problems most important incentive for the shows. From the occurring on a daily basis in the establishment. business owners’ perspective, appearing on the These businesses are not failing because they can- show may be a last-ditch effort to survive. Cer- not handle high volumes of customers; rather, the tainly, they expect a positive publicity effect from lack of customers is the problem!16 appearing on the show; and there is little to lose. Based on the legal court documents, published Unfortunately, the show hosts’ incentives are first-hand accounts from owners or employees, more aligned with creating an entertaining TV and our own personal off-the-record interviews show than in implementing the specific changes of owners and employees, it is clear that, to some that might have helped the business survive. degree, all of the shows use selective editing, scripted or enticed performances, hired actors, VI. CONCLUSION postfilming voice overs, and staged problems to increase the drama and “shock” factor for TV We have provided the first comprehensive viewers.17 Problems are made to appear worse analysis of the almost dozen shows that claim or more substantial than is actually the case to save failing businesses. Our analysis includes and shows focus on exacerbating interpersonal compiling data on every business that has appeared on the shows, examining customer reviews, conducting phone interviews with busi- 16. A good example where the owner tries repeatedly, and ness owners, and compiling first-hand accounts unsuccessfully, to stress this to the host during the episode is of owners, employees, and local reporters. the Windsor 75 episode of Restaurant: Impossible. The most important conclusion is that these 17. See, for example, Dehnart (2009), Cieslak (2016), Harrington (2012), McGee (2013), Jordan (2013), Wood shows themselves are businesses—reality TV (2014), Odam (2016), Murphy (2012), Reser (2016), Red- drama series that are influenced by their own dit (2013), Morabito (2010), and Fantozzi (2016). Also see incentives to be profitable and continue on air owner’s response to customer Michele L.’s Yelp posting on May 31, 2016 (Yelp 2016). In some cases, lawsuits provide through increasing viewership and satisfying details of these discrepancies. For example, see Hines (2007) sponsors through highly visible product place- and Wilkes et al. (2014). ments in the episodes. Certainly, personnel SOBEL ET AL.: CAN EXPERT TV HOSTS SAVE FAILING BUSINESSES? 23 training and publicity effects can have short-term slot on network TV and generate profits for the benefits to businesses profiled on these shows. hosts, crew, and production company—rather However, we find that the problems portrayed than saving many of the businesses they claim to are not always representative of the true issues be helping. facing the businesses. Indeed, some real-world major issues are entirely neglected, while other REFERENCES problems are manufactured to increase the drama Bar Rescue Updates. “Bar Rescue—Sorties Tavern for TV by editing, suggestive acting, and staging (O’Banion’s) Update.” 2014. Accessed August 26, unrealistic stress tests. In some cases, shows 2016. http://www.barrescueupdates.com/2014/03/bar- rescue-sorties-tavern-obanions.html?m=1. seek out and approach the businesses, again Bar Rescue Updates. “Bar Rescue—Second Base Bar and with an eye toward creating drama that makes Grill Update.” 2015. Accessed August 26, 2016. http:// for entertaining TV viewing. The solutions they www.barrescueupdates.com/2015/02/bar-rescue- second-base-bar-and-grill.html?m=1. implement are not always the solutions best for Bizzozero, P., R. Flepp, and E. Franck. “The Importance of the businesses’ survival, but rather create the best Suspense and Surprise in Entertainment Demand: Evi- placements for show sponsors or build the hosts’ dence from Wimbledon.” Journal of Economic Behav- personal brand or career. Lengthy legal contracts ior & Organization, 130, 2016, 47–63. Blumenthal, H. J., and O. R. Goodenough. This Business often prevent owners from discussing the show of Television: The Standard Guide to the Television and their experiences, and several shows have Industry. 4th ed. New York: Billboard Books, 2006. faced lawsuits over their portrayals. Viewership Bricker, T. “Gordon Ramsay Has Decided to End Kitchen Nightmares—Find Out Why.” 2014. Accessed is negatively correlated with the survival rate of August 26, 2016. http://www.eonline.com/news/ the businesses featured on the show, providing 553588/gordon-ramsay-has-decided-to-end-kitchen- little incentive for shows to assess their effective- nightmares-find-out-why. Chevalier, J. A., and D. Mayzlin. “The Effect of Word of ness or care about the businesses’ survival rates Mouth on Sales: Online Book Reviews.” Journal of after the interventions. There is no evidence the Marketing Research, 43(3), 2006, 345–54. shows have lower failure rates in later seasons as Cieslak, E. “A Week in Hell: Chelan Business Owners Dish a result of self-assessment. on Filming Upcoming Episode of ‘Hotel Hell’.” 2016. Accessed August 26, 2016 http://www.yakimaherald One show, The Profit, has a notable differ- .com/news/local/a-week-in-hell-chelan-business- ence in incentive structure as the host personally owners-dish-on-filming/article_8db86458-373d-11e6- invests in the business and continues involvement 9d87-83481d0af732.html. Dehnart, A. “Rachel Zoe’s Taylor Fired or Quit; Tabatha’s into the future. As theory would predict, this show Salon Takeover Made Salon Messier.” 2009. Accessed does have the lowest failure rate among major March 21, 2018. https://www.realityblurred.com/ shows, and data suggests these businesses may realitytv/2009/11/bravo-taylor_out_salon_reality/ Doyle, A. “Reality Television and Policing: The Case of not have been as close to failing as the ones fea- Cops,” in Arresting Images: Crime and Policing in tured on other shows. The businesses are selected Front of the Television Camera. Toronto: University of because they are good investments and the long- Toronto Press, 2003, 32–63. run survival and profitability of the business is Dunn, M. “Not Jus’ Donuts Featured on TLC’s Buddy’s Bak- ery Rescue.” 2014. Accessed August 26, 2016. http:// clearly in the host’s personal interest. www.houstonpress.com/restaurants/not-jus-donuts- Our research strongly suggests that an incen- featured-on-tlcs-buddys-bakery-rescue-6431860. tive differential plagues the effectiveness of “save Fantozzi, M. “Winter Haven Hotel Owner: Reality TV Show ‘Hotel Impossible’ Is Lying.” 2016. Accessed August the business” reality TV shows. At best they may 26, 2016 http://www.theledger.com/article/20160511/ be last-ditch efforts to generate temporary bumps NEWS/160519861. in revenue (or potential buyers) from the public- Fothergill, E., J. Guo, L. Howard, J. C. Kerns, N. D. Knuth, ity of being on TV for many of the businesses R. Brychta, K. Y. Chen, M. C. Skarulis, M. Walter, P. J. Walter, and K. D. Hall. “Persistent Metabolic Adap- featured. These same problems can be factors tation 6 Years after ‘The Biggest Loser’ Competition.” that limit the effectiveness of real-world business Obesity, 24, 2016, 1612–19. https://doi.org/10.1002/ consultants, local convention and visitor bureaus, oby.21538. Foxnews.com. “‘Fixer Upper’ House-Hunting Scenes Are or even the Small Business Administration. Fake, Show Participant Claims.” 2016. Accessed Jan- Whether it is a reality TV host or a hired private uary 4, 2017. http://www.foxnews.com/entertainment/ sector consultant that endeavors to help a strug- 2016/11/04/fixer-upper-house-hunting-scenes-are- fake-show-participant-claims.html. gling business, a differing incentive structure can Gardner, E. “Fired ‘Storage Wars’ Star Wins Round in Rig- result in an outcome that is less than successful. ging Lawsuit.” 2013. Accessed January 4, 2017. http:// Due to the strong financial incentives created by www.hollywoodreporter.com/thr-esq/fired-storage- paid sponsor product placements and the need wars-star-wins-619655. Harrington, D. “‘Hotel Impossible’ at Gurney’s Inn: Non- for drama that increases viewership, it would Reality Reality TV!” 2012. Accessed August 26, 2016. appear that, at best, the shows “save” their own http://www.hamptons.com/mobile//Hamptons-Style/ 24 CONTEMPORARY ECONOMIC POLICY

Main-Articles/17020/Hotel-Impossible-At-Gurneys- Reddit. “[AMA Request] An Employee of a Bar That Has Inn-Non-Reality.html#.V5kS9fkrJD9. Been Rescued on Jon Taffer’s SpikeTV Program, ‘Bar Hines, N. “Gordon Ramsay’s Lawsuit Nightmare.” The Rescue’.” 2013. Accessed August 26, 2016. https:// Times, June 20, 2007. Accessed August 26, 2016. www.reddit.com/r/IAmA/comments/1almg3/ama_ https://www.thetimes.co.uk/article/gordon-ramsays- request_an_employee_of_a_bar_that_has_been/. lawsuit-nightmare-rmj6pjb5h80 Reser, B. “The Truth about a ‘Bar Rescue’ Experience.” 2016. Interrobang. “Bar Rescue Miscues: Bars That Taffer Accessed August 26, 2016. http://brentreser.com/2016/ Couldn’t Save.” 2014. Accessed August 26, 2016. 04/the-truth-about-bar-rescue/. http://theinterrobang.com/bar-rescue-fail-john-taffer/. Ryland, A. “Kitchen Nightmares Lawyers Threaten Restau- Jordan, C. “Motel Owners Feel ‘Punk’d’ by ‘Hotel Impos- rant Owners: Shut Up—or You Owe Us $100K.” sible’.” 2013. Accessed August 26, 2016. http://www 2013. Accessed August 26, 2016. http://radaronline .usatoday.com/story/news/nation/2013/10/25/hotel- .com/exclusives/2013/05/kitchen-nightmares-lawyers- impossible-thunderbird-motel/3200815/. threaten-restaurant-owners-amys-baking-company/. Judge, G., G. Carter, R. Hill, W. E. Griffiths, H. Lütkepohl, Satran, J. “Most Restaurants ‘Saved’ by Gordon Ram- and T. Lee. Introduction to the Theory and Practice of say’s ‘Kitchen Nightmares’ Are Now Closed.” 2015. Econometrics. 2nd. ed. New York: John Wiley & Sons, Accessed August 26, 2016 http://www.huffingtonpost 1988. .com/2015/06/24/kitchen-nightmares-restaurants_n_ Katzenbach, J., I. Steffen, and C. Kronley. “Cultural Change 7649462.html. That Sticks: Start with What’s Already Working.” Schneider, B. “The People Make the Place.” Personnel Psy- 2012. Accessed January 5, 2017. https://hbr.org/2012/ chology, 40(3), 1987, 437–53. 07/cultural-change-that-sticks. Schumpeter, J. A. The Theory of Economic Develop- Klein, M. “ ‘Bar Rescue’ Clients Give Up on Taffer’s ment. Cambridge: Harvard University Press, 1934 Changes.” 2014. Accessed August 26, 2016. http:// [1911]. www.philly.com/philly/blogs/the-insider/Taffer-fails- Segal, D. “After the Cameras Leave the Kitchen.” 2012. Lickety-Plush-Bar-Rescue-Villaris.html. Accessed August 26, 2016 http://www.nytimes.com/ Lawson, R. A., and K. Lawson. “Good Cops, Bad Cops, 2012/07/11/dining/what-happens-after-restaurant- Whatcha Gonna Do?” Journal of Private Enterprise, impossible-leaves-town.html?_r=0. 31(4), 2016, 85–96. Sobel, R. S. “Economic Freedom and Entrepreneurship,” in Lore, D. C. “Staten Island’s Not Just Bagels to Be Featured What America’s Decline in Economic Freedom Means in Fall Reality Series.” 2013. Accessed August 26, for Entrepreneurship and Prosperity, edited by D. J. 2016.http://www.silive.com/eastshore/index.ssf/2013/ Boudreaux. Vancouver, BC: The Fraser Institute, 2015, 07/staten_islands_not_just_bagels.html. 37–66. Magder, T. “The End of TV 101: Reality Programs, Formats, Sobel, R. S., J. R. Clark, and D. R. Lee. “Freedom, Barriers and the New Business of Television,” in Reality TV: to Entry, Entrepreneurship, and Economic Progress.” Remaking Television Culture, Chapter 7, edited by S. Review of Austrian Economics, 20(4), 2007, 221–36. Murray and L. Ouellette. New York: New York Univer- Taraborelli, J. “Farewell, DownCity: A Quirky and Enduring sity Press, 2004, 137–56. Downtown Eatery Closes.” 2011. Accessed August 26, McGee, J. “Chappy: ‘Kitchen Nightmares’ High on Stress, 2016. http://providenceonline.com/stories/DownCity- Low on Reality.” 2013. Accessed August 26, 2016. restaurant-closes-Providence,460. http://www.bizjournals.com/nashville/blog/2013/01/ The Profit Success. “Jacob Maarse Florists.” 2015. Accessed sfjdkjs.html?page=all. August 26, 2016. http://theprofitsuccess.blogspot.com/ Mogel, L. This Business of Broadcasting. New York: Bill- 2015/05/jacob-maarse-florists.html. board Books, 2004. Tomicki, H. “The Feeding Tube: How Much Are TV Appear- Morabito, G. “The Total Fakery of a Kitchen Nightmares ances Worth for Restaurants?” 2011. Accessed June 8, Taping.” 2010. Accessed August 26, 2016. http://www 2016. http://www.grubstreet.com/2011/10/restaurant- .eater.com/2010/5/7/6734309/the-total-fakery-of-a- tv-appearance-impact.html. kitchen-nightmares-taping. Tran, K. “Bardonia Business to Be Featured on Food Murphy, N. “Piratz Tavern: How Bar Rescue Faked Reality.” Network.” 2012. Accessed August 26, 2016. http:// 2012. Accessed August 26, 2016. http://k23detectives patch.com/new-york/nanuet/bardonia-business-to-be- .com/2012/09/23/piratz-tavern-bar-resue/. featured-on-food-network. Odam, M. “Satay Closing Thursday Night after Almost Walthall, J. W. “Did Gordon Ramsay Change El Greco for the 30 Years in Business.” 2016. Accessed August 26, Worse?” 2011. Accessed August 26, 2016. http://austin 2016 http://dining.blog.austin360.com/2016/06/15/ .eater.com/2011/10/3/6646849/did-gordon-ramsay- satay-closing-thursday-night-after-almost-30-years- change-el-greco-for-the-worse. in-business/. Weiss, K. J. “Hoarding, Hermitage, and the Law: Why We Offitzer, A. “Extreme Store Makeover: New Milford’s Wine Love the Collyer Brothers.” Journal of the Ameri- and Roses Is Ready to Bloom.” 2013. Accessed August can Academy of Psychiatry and the Law, 38(2), 2010, 26, 2016. http://www.northjersey.com/news/extreme- 251–57. store-makeover-new-milford-s-wine-and-roses-is- Wilkes, P. T., Bongo LLC, J. Taffer, and N. Taffer. “Supe- ready-to-bloom-1.715312. rior Court of the State of California for the County Oliver, M. B. “Portrayals of Crime, Race, and Aggression of Los Angeles.” 2014. Accessed August 26, 2016. in ‘Reality-Based’ Police Shows: A Content Analysis.” http://www-deadline-com.vimg.net/wp-content/ Journal of Broadcasting & Electronic Media, 28(2), uploads/2014/04/Bar-Rescue-assult-lawsuit-WM__ 1994, 179–92. 140407223510.pdf. Phu, L. “Hotel Impossible Wraps Up Filming at Alaskan Wood, V. B. “What If Your Restaurant Doesn’t Need Hotel.” 2013. Accessed August 26, 2016. http://www Redemption? Getting Satay Ready for Its Close- .alaskapublic.org/2013/09/30/hotel-impossible-wraps- Up.” 2014. Accessed August 26, 2016. http://www up-filming-at-alaskan-hotel/. .austinchronicle.com/food/2014-01-31/what-if-your- Platt, A. “The Diners, Drive-Ins and Dives Effect.” 2015. restaurant-doesnt-need-redemption/. Accessed June 8, 2016. http://tcbmag.com/Industries/ Yelp. “The Canyon Inn.” 2016. Accessed August 26, 2016. Food/The-Diners-Drive-Ins-And-Dives-Effect. http://www.yelp.com/biz/the-canyon-inn-yorba-linda.