Abstract Relying on qualitative interviews and thematic analysis, this paper explores how transnational call centers in India evolve practices in an attempt to counter the occasional hostility by customers based in the United States. These range from techniques which prevent customers from recognizing that their call is being routed to India through the use of neutralized accents, 'locational masking' and familiarizing agents with American culture to Taylorist modes of ensuring 'passivity' in the agent in the face of a hostile customer. I argue that these practices form an inherent part of transnational call center work and subsequently lend to its uniqueness as a poster child of globalization in a brave new post industrial world.
"When we listened back to calls people had complained about often they were fine. Some people wanted the member of staff to fail because they were in India. I don't know why that should be, but when customers start voting with their feet, you have to respond."
-Adrian Web (Call Center Manager, BBC news, 2/14/2007) quoted in Wang (2009).
Introduction
Globalized work processes have fundamentally altered the ways in which labor markets are organized around the world (Mirchandani 2004). The global outsourcing industry also referred to as the Business Process Outsourcing (BPO) industry is arguably the poster child of globalization. It is a relatively new industry that has been growing at a very fast pace in the last few years around the world (Deery & Kinnie 2004). It includes a range of services, from back office operations such as airlines ticket and insurance claims processing and medical transcription services, to call centers that provide 24/7 customer support, back office services, telemarketing, for banks, credit card companies, computer companies and the like (Upadhya 2006). The biggest trend in this industry in recent times has been a move to offshore work to countries like India and the Philippines where there is a large availability of cheap, but high quality, English-speaking labor - the key ingredient for running call centers (Dossani & Kenney 2003). India being a former British colony possesses a large pool of English-speaking workforce willing to work at salaries which are a fraction of those in the OECD countries. Since, offshoring is structured to take advantage of labor arbitrage arising out of wage differentials between countries; India is uniquely suited to be an offshoring destination. About 80 percent of the call center industry in India caters to international markets (Holman, Batt, and Holtgrewe 2007) namely those of the US, UK and Australia. In fact, this industry in India was entirely a product of globalization since there was no domestic call center industry catering to the Indian market at the time when the international call centers started (Sitt 1997).
In the last decade, the BPO industry has grown exponentially leading to the large scale creation of entry-level white collar jobs in India and the consequent loss of jobs in the OECD countries. The popular sentiment in first world countries like the United States has been largely anti outsourcing with the subsequent politicization of the issue at the national level with the 2005 Democratic Presidential nominee John Kerry's infamous declaration on 'Benedict Arnold' companies' ("'Benedict Arnold' CEOs," The Wall Street Journal, February 12, 2004.). One UK MP predicted that 'there would not be a call centre in Scotland in five years (Brian Donoghue. Herald, 14 March 2003, cited in Taylor and Bain 2005).
The BPO industry in India can be divided into two broad categories - voice-based processes and non voice-based processes. Call centers or contact centers are a component of the BPO industry, relating to voice-based processes. The call center industry is well situated within India's global leadership with its offshore information technology and business process outsourcing industries increasing at an annual rate greater than 25% and generating export revenues of $60 billion by 2010 (NASSCOM, National Association of Software and Service Companies, 2005, cited by Pal & Buzzanell 2008). As per the Holman report (Holman et al. 2007), a typical Indian call center has about 500 employees, is about four years old, services the international market (80% international customers), and focuses on outbound calling (60% revenue from outbound sales).
Indian call center work involves employees ('agents') providing voice-to-voice service to clients primarily in North America and the United Kingdom. These services are mostly the "high-volume, low-value, routinized" end of call centre work which tends to be shifted to India (Taylor and Bain 2005). Agents are required to learn American accents, work at night to cater to U.S. time zones, and adjust to an altered social and family life. They are expected to be conversant with day-to-day American issues to the extent that they are able to carry on casual conversations with clients (Mirchandani 2004).
At the macro-level, the emergence of this industry also represents a significant shift in globalization. Unlike silicon chip production in Taiwan, maquiladoras in Mexico, or McDonald's in France, transnational customer service employment represents a shift from exporting the production of material goods or culture to a full scale reproduction of identity and culture. In contrast to McDonalds selling french fries in Paris, but not requiring an American accent from its French employees, call center operations are based on the availability of workers trained to embody an American identity and cultural cues (Patel 2006).
Methods
The primary purpose of this paper is to explore the nature of transnational call center work within the context of customer hostility to outsourcing and the practices adopted by the call center management to negotiate this hostility. At one level, these practices are preventive in nature i.e. prevent the customer from realizing that their call is being routed to India. When that doesn't succeed, call centers use Taylorist and 'panoptical' control mechanisms over agents in an attempt to ensure that they remain 'passive' on the face of hostility, which (hopefully) prevents the situation from getting aggravated. A review of existing literature on call centers and field work has been used to analyze this phenomenon. Semi-structured interviews were conducted from June to November 2009 in a BPO company in Gurgaon which is in the suburbs of New Delhi. I shall call it ABC International (not its real name). ABC International, incorporated in 2007, is a 78,561 square feet facility spread over two floors employing over 600 people as on November 2009. Out of 600 employees, around 440 are men and 160 women. It has no domestic clients and exclusively serves the international market. Its business lies largely in the domain of banking and financial services.
The participants can be categorized in three broad classifications- the agents i.e. the customer representatives, their trainers and the managers. Most of the agents are in their early or mid twenties with the managers being in mid twenties to early thirties. I conducted semi-structured interviews with nineteen agents, two trainers and one manager. The agents were randomly selected and can be broadly considered as representative of call center agents in Gurgaon. In terms of the hierarchical structure, the agents were the junior most and were designated as Band 1 or Band 2. Eleven agents were male and eight female. Three male agents were married, though none was married more than eighteen months. Only one female agent was married for less than six months. All in all, they confirmed the general perception of a 'youthful workforce'. They reported to their team leaders who were either Band 3 or 4. The two trainers- a male and a female were involved in providing voice and accent training to the agents were respectively Band 3 and 4. The manager was the senior most respondent at Band 5 and supervised the two trainers as well as the team leaders.
The agents dealt with both incoming as well as outgoing calls. The incoming calls were mostly customers dialing a 1-800 (i.e. a U.S. toll free) number to resolve problems they faced with products like credit cards. The outgoing calls were a part of a process termed 'Collections'. Collections involve recovery of overdue payments from the customers. They earned between USD 290 to 390 per month depending on performance and experience. Nearly one-fourth to one-third of the compensation was linked to performance. For outbound callers, the defining criteria was achievement of collection targets, while for inbound callers, the customer satisfaction ('C-Sat') ratings were used to determine bonus payouts.
While most agent interviews were one-to-one, in four cases other call center workers were present. However care was taken to ensure that no team leader or manager was present. All except for one took place in or around the office premises like cafeteria, meeting rooms etc. A disclosure is necessary: I was also employed in the same company as an assistant manager since May 2008 albeit in a different division which had no direct link with the call center operations. None of the interviewees were from my unit and I made it clear that the purpose of my questions was strictly academic. I also assured them of the confidentiality of their responses and made sure they were comfortable conversing with me before I started the interview. Nobody who was even marginally hesitant was interviewed or was allowed to observe the interviews. All in all, I have no reason to believe that the bias in their responses was any greater than that those which might have been elicited by an external interviewer. Also being employed in a BPO for a year and a half provided ethnographic opportunities unavailable to an external researcher.
'Neutralized accents' and 'locational masking'
British and American media have featured stories of agents in Indian call centers modifying their accents in conversations with customers in Britain and the USA. These stories have had wide interest and appeal for two reasons. One is the novelty of speaking to someone in India about a product purchased in Europe or the US, thanks to the way that technology has enabled the outsourcing of the service industry in recent years. The other reason for the publicity is the job losses in the West associated with this development. These local effects have meant that many global workers have faced hostility from customers (Cowie 2007).
ABC International requires agents to assume Western identities in providing service. Agents also undergo training in neutralized accents [1] and are discouraged from disclosing their geographical location i.e. India on the phone. This appears to be a fairly common practice as indicated by a review of literature on Indian call centers. Mirchandani (2003) refers to this phenomenon as 'locational masking' while Poster (2007) calls it 'national identity management'. Many call centers have non-disclosure agreements with their clients that require them to develop protocols through which their location in India is not revealed to customers (D'Cruz and Noronha 2004). Consequently, agents try to mislead customers about their location or divulge only the bare minimum. Accordingly, agents are trained to avoid answering questions from customers about their location or reveal their nationality. For instance, D'Cruz and Noronha (2004) say that they were trained to say that they were headquartered in the United States or United Kingdom or any other country as the case would be. When quizzed further, they either refused to disclose any more information about their location citing security reasons, or they would mention that they were located in Asia. Only if the customer persisted would they give in or allow the customer to hang up (D'Cruz and Noronha 2004).
The rationale for these practices, according to the manager in ABC International, is that they allow agents to serve the customer better. Revealing that service work has been subcontracted to India may give rise to customer dissatisfaction for a wide variety of reasons, ranging from racism and ignorance towards Indians to assumptions about exploitative transnational corporate practices. These concerns are not unfounded; workers report that they frequently face racism in their interactions with customers (Mirchandani 2004). But another reason, less discussed, is to mute the political backlash in the West about the morality of offshore outsourcing (Nadeem 2009). Nadeem (2009) also draws a parallel between the need to don Western pseudonyms and accents with Macaulay's image of a class of pliable intermediaries adopting the culture and language of the colonial power [2] .
Indian accents do more than turn the occasional customer hostile; they are a major determinant of customer satisfaction as well. In an experiment conducted by Wang, Arndt, Singh and Biernat (2009), accent stereotypes were demonstrated to influence both customer evaluation of employee performance and customer satisfaction. American customers were found to be more likely to attribute favorable service to individual employee factors and attribute unfavorable service to external firm factors when an employee has an American or British accent. Even when service scenarios and conversation scripts were identical, customers rated the performance of employees with Indian accents much lower than the performance of employees with American and British accents. Further, when employees from a favorable group (British or American) performed well, customers attributed this favorable outcome to internal factors. When they provided poor service, customers attributed this unfavorable outcome to external factors. The reverse is true when the employee was from an unfavorable group i.e. Indian. Not surprisingly, disguising nationality is a frequent management practice in Indian call centers. While I am not aware of the controls used by Wang et al as to whether the Indian accents materially impacted the subject's comprehension, I am inclined to believe that the reduced customer satisfaction is not caused by their inability to comprehend the Indian accents per se but by the belief that the call was being routed to India, i.e. the anti-outsourcing sentiment.
Cultural familiarization
Attempts to modify natural speech are further combined with programs designed to familiarize agents with the cultural background of customers' countries. Cultural familiarization is an extension of locational masking reducing the chances of the identification of the agent as non-American. Taylor and Bain (2005) recorded a centre providing two-day modules on US and UK culture, covered 'typical customers, geography, currency, holidays, time-zones, slang and colloquialisms, and major sporting events'. ABC International also has a training module on cultural familiarization. The lifestyle of the people including standards of living, economic and political conditions, sports and sports stars, entertainment industry which apart from Hollywood films includes popular sitcoms and shows, education system and popular brands of common products. Training methods involve videos, playing recordings of calls and also scenes from sitcoms and films.
However, cultural familiarization is frequently problematic, as Mirchandani (2003) has pointed out. Whether the objective is 'making Americans', or making English, Scots, or whatever nationality Indians are expected, chameleon like to become, there are limitations on the ability of these neo-colonialist practices to succeed (Mirchandani 2003). A male trainer in ABC International, recounting his experiences of facing some resistance from agents whom he was training to 'become an American', gave this pep talk to his team:
This is your choice…you have chosen to work here. It's a part of your job to sound like an American, so I would like to see some enthusiasm now.
(Trainer, ABC International)
One 22 year old female agent described these attempts by trainers at 'to make her an American' while on calls theatrical (or 'drama' as she put it). Another 26 year old male agent felt that trying to sound American felt humiliating:
You are made to feel (as if) being an Indian is somehow inferior to being an American…Agreed they are very rich but it doesn't mean we have to stoop to them.
(Agent, male)
In ABC International there seemed to be a disagreement as the extent of the utility of these practices between the agents and their trainers. While both trainers felt that it worked 70-80% of the time, out of the 19 agents I interviewed, 16 felt that the customer was able to 'see through' the guise at least 50% of the time and therefore considered this practice to be unnecessary and stressful:
You get to know when the customer realizes that I am not an American. Subtle changes in the tone and voice indicate this. Sometimes they become condescending and sarcastic; at other times outright hostile and ask for the call to be transferred to an American.
(Agent, female)
"Can I speak to an American?" or "Can I speak to someone who speaks in English?" are the first things you hear when you start working in this job. Americans are not as dumb as we think… they get to know most of the times that we are faking accents. While most of them remain polite some become really aggressive.
(Agent, male)
I recall an incident when a customer said "You can't fool me…. You aren't speaking from Florida, I know you are from India, you speak like an Indian. I want this call transferred to America."
(Agent, male)
Westernized pseudonyms, locational masking and attempts at cultural familiarization add significant, culturally-specific dimensions of psychological strain to pressures associated with the emotional labor that call centers engender (Taylor and Bain 2005). Furthermore, the synchronization of agents' shifts with western customer servicing hours is both symbolic of the subordination of Indian workers to the interests of western capital and, because it induces nocturnal work patterns, significantly exacerbates the generic pressures associated with routinized call-handling (Taylor and Bain 2005). Taylor and Bain (2005) have gone on to describe these practices as neo-imperialist and indeed racist. They have further argued that it cannot be assumed that globalization goes unchallenged and that Indians become eager agents - in both literal and metaphorical senses - willing to re-model themselves at the behest of American or British corporate clients.
While the degree of success of these practices is debatable, it is unanimously agreed that they are not universally successful. This necessitates the adoption of preventive measures to contain the subsequent hostility shown by some customers. While there seems to a divergence of opinions between the agents and managers on the extent of this hostility, there is no doubt that it exists and is fairly pervasive. [3] Since the author hasn't spoken to the customers, the extent to which this can be genuinely attributed to their inability to comprehend Indian accents (or a certain 'irritation' with affected accents [4] ) is difficult to say.
The challenge posed for call centre managements was to prevent the agent from responding to such hostility in an aggressive manner such as 'talking back' or abruptly terminating the call. What management required was a mechanism to maintain control over the agents. Enter Taylorism.
'Taylorism': Back from the dead
Some eighty years after being discredited in the Hawthorne experiments Taylorism thrives in Indian call centers. The types of calls handled tend to be highly standardized, simple in content, tightly scripted and of short duration (Taylor and Bain 2005). While Taylorist terminology like time and motion studies are not used, their call center equivalents like average handling time (AHT) and automatic call distribution (ACD) system are used liberally. The monitoring of the agents on the floor using sophisticated technology appears to be an equivalent of the much reviled Taylorist foreman armed with a stopwatch. Standardized scripts are yet another reflection of the Taylorist belief in "the one best way" to do a job. More than anything else is the surprising lack of participative management practices in the day to day running of ABC International. The service level agreements with clients lay down detailed descriptions of the work flow which are then followed to the letter and no deviations are permitted without the express approval of the client. It is this enforced standardization which reveals the extent to which call center processes are based on Taylorist principles [5] :
The entire workflow is automated and monitored. It has to be…there is no other way to ensure that an agent handles calls properly.
(Manager, ABC International)
Upadhya (2006) describes the extent of Taylorism prevalent in a Bangalore call center, through the use of computerized surveillance and monitoring systems, which according to her, represents rationalization of the work process taken to an extreme. She describes the work process to be controlled through very exact, rigid, and 'panoptical' systems of monitoring, measurement, and management, using sophisticated computer and telecommunications systems that bind the human worker tightly to the machine. The organization of BPOs there is almost fully automated and controlled through sophisticated computer systems and telecom networks that tie the workers closely to their machines and to the entire system. Most BPOs also have closed-circuit video cameras scattered around the floor to provide additional 'panoptical' capacity. Team leaders and other managers also listen in on calls periodically to keep a check on workers' performance. Performance criteria ('metrics') are project or customer-specific and determined by the contract that the service provider enters into with the client, known as a service level agreement (SLA). Metrics include both quantitative criteria like average handling time (AHT) or number of calls handled per shift, percentage of sales closed, number of new members registered and so on. In addition to these quantitative measures, the quality of work performed is tracked and evaluated, and these assessments are fed into the performance appraisal process. Samples of calls are also regularly recorded to be evaluated later by trainers and quality personnel. Recording of calls enables managers and trainers to play back calls during coaching sessions and point out problems to the agent, and also provides proof of wrongdoing or mistakes when the worker is being given a 'corrective' (Upadhya 2006). Standardized service scripts are another of the central mechanisms to control the nature, timing, norms and structure of work (Mirchandani 2004).
ABC International is also organized on nearly identical lines regarding Taylorism, surveillance and standardized scripts. That the above description of a Bangalore-based call center (Upadhya 2006) fits ABC International which is based in Gurgaon (which is at a distance of 1280 miles from Bangalore) so perfectly shows the extent of uniformity of practices and the pervasiveness of Taylorism in transnational call centers in India. The stated reason by management is the compliance with the service level agreement. The manager who was a male in his early thirties had been working in the call center industry since 1998, observed the extent of surveillance and emphasis on standardized scripts had actually increased over the last ten years. To some extent, this was due to the increased availability of 'surveillance technology'. In ABC International as per policy, all calls are recorded and preserved in an encrypted form (so as to secure confidential data) until the client expressly requests for their destruction.
My question is whether the need to maintain a non-aggravating tone in the face of customer hostility is a determinant of the extent of Taylorism prevalent in call centers. My argument is that while no overt mention is made of customer hostility arising out of anti-outsourcing sentiments, racism etc., in the service level agreements, one of the main reasons for BPOs to insist that customer calls be handled as per the script is to ensure the 'passivity' of an agent in the face of such hostility. Upadhya (2006) mentions that agents find abusive calls (which are usually anti-Indian and linked to the anti-outsourcing feeling in the U.S.) as particularly galling, and some find ways of 'getting back' at the customer for abusing them or insulting Indians. One perceptive male agent in ABC International, who was in his mid-twenties, felt that night shifts and abusive customers was the main reason for the 'stigmatization' of call center jobs. A female agent recalled breaking down while handling the call of one particularly difficult customer. The manager informed me that while he did not recall any agent citing customer hostility specifically as a reason for quitting, he felt that it invariably contributed to the 'burnout' process.
The call center managements anticipate some level of customer hostility and consequently take steps to prevent the agents from aggravating the issue. Talking back in the face of customer hostility is considered a 'breach' by the agent. Call center management can get alerted of a breach if an agent is caught during call monitoring or due to a complaint from a customer.
If an agent gets caught terminating the call abruptly of 'talking back' when facing customer hostility, the first thing we do is to take him off the floor. We then review a sample of calls handled by him in the past and decide what should be done with him. The action taken by the company may range from putting him back on training to termination of services.
(Manager, ABC International)
In case the customer resorted to profanity, ABC International had a policy of requiring the agent politely warning the customer thrice that he would disconnect the call unless the customer desisted from profanity. After three warnings the agent was permitted to disconnect the call. However, in practice the call is seldom disconnected even in cases when it would have been permissible to do so. The reason that the agents gave was that they could earn extra 'points' if they succeeded in pacifying an abusive customer. Such points would be reflected in their monthly appraisals and could result an increase in their variable compensation. Also, all disconnected calls are required to be reviewed by the team leaders. In case the disconnection was perceived as unjustified, it was considered as a breach. Thus, while on paper the management permits the termination of a profanity-laced call, it tacitly discourages agents from doing so. Mirchandani (2008) argues that the behavior of abusive callers is implicitly sanctioned, given that the workers must provide three warnings before disconnecting a call. It is only repeat abusers who can be challenged (Mirchandani 2008).
While the manager whom I interviewed did not wish to place on record the reason for such a practice, one possible explanation can be the local labor laws requiring a conducive work environment. A company policy insisting on profane calls being taken might run afoul of the local labor laws.
Conclusion
The preponderance of neutralized accents, 'locational masking' and cultural familiarization are unique to the call center industry having no parallel in any other industry. They are also pervasive through out India- indeed I was quite amazed at the extent to which practices at ABC International were nearly identical to other call centers as described by other researchers. This pervasiveness appears to suggest the inevitability of applying these practices when running the operations of a transnational call center industry in India. Additionally, the responses provided by the manager and the two trainers of ABC International strengthen this perspective. While, these practices are sought to be justified by need to counter the anti-outsourcing sentiment prevalent in the United States, they, in my opinion add a new dimension to an existing critical narrative of call centers which is currently centered on critiquing the neo-imperialist process of reducing young, educated Indian workers to 'cyber-coolies' (Mirchandani 2004) or in other words, a low-skilled workforce as the twenty first century version of Macaulay's army of clerks.
Further, the 'passivity' of the agent in the face of rabid customer hostility raises a few questions about its prevalence. Questions arise as to the reason agents continue to remain passive in what is unquestionably an unpleasant experience. While to some extent it is enforced by the management by the use of the Taylorist and 'panoptical' practices, it is undeniable that the agents are frequently willful fait-accompli, virtually amounting to an ethic of subservience. Further evidence comes from the fact that, agents continue to try placating an abusive caller, even when they are permitted to terminate the call. At some level, this ethic of subservience appears to arise out of a certain level of subalternity of the workforce, who, while being graduates, mostly come from the lower middle class and therefore are not 'classy' enough (or don't have the requisite cultural capital) to be able to compete for the high end white-collar jobs and are thus resigned to working in a call center. However, call center employment also allows them to access to elite spaces like swank offices in high-rise steel and glass buildings, American accents, relatively higher disposable incomes and an exposure to American culture, not to mention salaries they never dreamed they could earn. All of this is highly aspirational to the middle and lower middle classes, to which this workforce mainly belongs. It is these trappings which create the notion of a call center work as a privileged occupation. Mirchandani (2004) also notes the role of the extensive screening process in place at the recruitment stage in furthering this impression. Possibly, the consequent gratefulness that the agent feels in being employed in a 'privileged occupation' outweighs the inherently demeaning and subservient task of handling abusive callers.
Secondly, as a conjecture, it could be suggested this passivity may also arise from invoking the Indian ethic of hospitality which advocates the primacy of the guest viz. treating the American caller as the guest and the agent, his host, thereby binding the agent to a degree of politeness virtually amounting to subservience. I would like to examine this idea in greater detail in my future papers and hopefully draw a comparison with call centers in other Asian countries which have a similar ethic of hospitality. Interestingly, Mirchandani (2004) notes that through their discussions of their customers, workers construct Americans as 'rich but stupid', thereby allowing workers to pity rather than revere their American customers. This latter conception comes from the equally prevalent notion of India as a budding economic superpower which does not need to bow to anyone and this leads to a productive tension in the workspace and a sense of identity and self.
Then there are a few other questions which I would like to examine in the near future. Given the conservative and patriarchal nature of Indian society, it would of interest to investigate how sexual abuse, if any, is negotiated by female agents. Also, are there any gender differences in terms of negotiating hostility? Further, I would like to examine the negotiation of hostility by the marginalized groups like the like lower castes, and religious minorities like Muslims and indigenous peoples who are at the receiving end of all sorts of prejudice in their local contexts. Are they better at handling hostility vis-à-vis upper caste Hindus given sustained histories of oppression? In other words are they more tolerant of an abusive customer? Also, how do they negotiate the radicalized inflection of abuse they might have to face e.g. racism from US clients?
The uniqueness of the transnational call center industry in India becomes clearer when one considers the fact that to some extent, their success depends on its customers not realizing that it exists (i.e. their call is being routed to India). While their practices appear to be neo-imperialist, one must however acknowledge that it offers previously unavailable employment opportunities to India's burgeoning youth. As Mirchandani (2003) argues, highlighting the continually contested and heterogeneous nature of global capitalism would allow for further analysis of the micro processes within which resistance to transnational subcontracted work occurs as the industry develops in India. Globalization does not exist in some rarefied stratosphere. It always touches ground (Brah 2002). Transnational call centers in India offer an opportunity to study one such 'touch down'.
