The value of hiring through employee referrals in developed countries

Turnover

One important pattern for referred employees is that they tend to have lower turnover rates than non-referred employees. This pattern has been documented in various labor market settings, including the US [2], Germany [3], and online labor markets [4]. In the US, lower turnover has been found in representative labor market studies [2], as well as in the context of certain jobs, including call-center workers, truckers, workers in financial services, and high-tech workers [5], [6].

One interesting finding is that differences between referred and non-referred employees in turnover may be larger at the beginning of a new employee’s tenure [3]. This gap may fade as tenure increases. Thus, after two employees (one referred and one non-referred) have been at a company long enough, the relative benefit in terms of turnover risk may start to diminish, though the literature does not clarify what qualifies as “long enough”. One possible explanation for this is that referrals are initially better-matched with firms than non-referrals. As information about fit between employees and firms is gradually revealed, the difference in fit between referrals and non-referrals diminishes over time, meaning that differences in turnover risk also diminish over time.

Conversely, other studies do not find that the turnover gap shrinks with tenure [6]. Differences in findings could reflect differences in the environment being studied; for example, one paper analyzes data from a large European metropolitan market [3], whereas the other studies data from a large US financial institution [6]. The findings in the European study may reflect trends in the labor market at large, whereas there may be variability within particular firms [6].

Productivity

Besides turnover, another very important variable for firms is worker productivity. It is thus relevant to examine if and to what extent employee referrals lead to more productive workers. Unfortunately, relatively little information exists about referrals and productivity because worker-level productivity data sets are not readily available. As summarized in [4], study findings have run the gamut, from referred employees having higher productivity, to referred employees having lower productivity, to referrals and non-referrals having similar productivity on most metrics.

However, the same study argues that selection during the hiring process can play an important role in whether referred employees are more productive than non-referred ones [4]. Firms try to hire the best candidates during the selection process, and differences in the selection process between referrals and non-referrals may affect whether there are productivity differences between referrals and non-referrals. Using an online labor market, the study obtained a large number of referred and non-referred applicants. However, rather than trying to hire the best applicants, the firm hired all applicants. This served to eliminate the selection pressure that typically occurs when firms attempt to hire the best available candidates, which frequently ends up with firms favoring referred candidates. The study can therefore shed light on whether referred applicants tend to be more productive than non-referred applicants. The authors find that referrals tend to be more productive than non-referrals, but that selection during the hiring process may be important for understanding productivity comparisons in general. For managers, this raises the possibility that the way that referrals and non-referrals are selected during the hiring process affects the nature of productivity differences after hiring.

Less recruiting cost

Another way in which referrals could be valuable for firms is if they lower the costs that firms incur when hiring new workers. For instance, firms need to spend less time conducting interviews and reading resumes when relying on referrals. Various studies in sociology and economics therefore argue that referrals help reduce recruiting costs [5], [1], [7]. That is, to achieve a hiring outcome through an employee referral, the cost of recruitment may be lower than to achieve a hiring outcome for a non-referred individual.

Good homophily

“Homophily” is a term often used by sociologists. It refers to people’s tendency to associate with people like themselves. Homophily is observed across a wide range of social phenomena. For example, people tend to be friends with people of similar ethnicities to themselves.

One way that employee referrals can be beneficial for firms is if they help leverage homophily with respect to desirable characteristics, such as productivity, attrition risk, or overall fit for the firm. For example, one recent study found that truck drivers tended to refer people with similar productivity (either in terms of miles or accidents) [5]. Thus, provided that individuals with higher productivity are more likely to make referrals (which could result for different reasons such as high productivity workers being more committed to an organization), firms can potentially benefit from those employees’ likeminded (i.e. similarly productive) referees.

One study on referrals in online labor markets also found that homophily on productivity is a common feature of referral relationships [4]. That is, more productive employees tend to refer people who are also more productive.

Potentially disadvantaging women and minorities

Despite its positive aspects, referral hiring could present a challenge for firms’ workplace diversity, especially given the tendency for employees to refer candidates with similar demographics as themselves [6]. One recent study argues that referrals could decrease diversity in the workplace [8]. The authors used data from a field experiment in Malawi and found that women seemed to be disadvantaged by the referral process relative to men. Specifically, they found that men were unlikely to refer women, and that women referred applicants who tended to perform worse (in terms of qualifying for a job) than applicants referred by men. This is an important finding, and it would certainly be worthwhile to examine similar issues in developed country economies as well (as cultural norms related to gender may be different between developed and developing countries).

Turning to ethnicity and back to the developed world, another study provides evidence from the US that black youth may receive less of a benefit from using informal methods to acquire jobs than white youth [9]. This occurs even though whites and blacks do not significantly differ in their likelihood of using friends or relatives in the job-finding process. In contrast, a study from the mid-1990s did not find lower returns to referrals for blacks (focusing on wages instead of job-finding), but did find lower returns to referrals for Hispanics.

In a rich and detailed ethnography, a book published in the late 1990s shows that disadvantaged blacks and Hispanics in New York frequently use social networks to find jobs. However, these social networks may lead to lower-skilled work, such as fast-food jobs. Individuals may be disadvantaged relative to others if they have fewer individuals with higher-skill or higher-paying jobs in their networks. This suggests an important point for policymakers that it is not only the size/quantity of an individual’s network that affects what types of job that person learns about, but also the “quality” of a person’s network. Increasing referral-based hiring could potentially disadvantage certain groups of people if there are differences across groups in the nature of their job networks.

The interaction between job networks and characteristics such as ethnicity and gender is complex. In the sociology literature, an influential paper from the mid-2000s argues that reputational considerations play a large role in the referral process. Specifically, minorities may be hesitant to make referrals if they are concerned that they will suffer reputational consequences from a referral that works out poorly. If an employee believes that their own position in the workplace is tenuous or that they have few outside options to pursue if things go sour, then that employee may be reluctant to make a referral.

Higher wages

One limitation of hiring through referrals (from the firm’s perspective) is that firms may need to pay referred employees more compared to non-referred employees (e.g. if referrals have higher productivity or match quality, or if referrals have an inside edge in negotiating higher salaries). The evidence is mixed, however, as different studies have reached different conclusions about whether referred employees earn more than non-referred employees [2].

One recent study examined differences over time between referred and non-referred employees in terms of wages at a large financial institution [6]. It found that referred employees started off earning more than non-referred employees. Over time, however, that difference faded away. A similar set of findings was recovered using data from Germany [3]. This pattern may reflect employers obtaining more precise initial information about worker quality for referred employees compared to non-referred employees. It should be noted that initial wages for referrals and non-referrals are the result of a complex hiring process. Moreover, paying higher wages is not necessarily a disadvantage for the firm if the higher wages reflect an employer having lower uncertainty about worker quality, as is the case of referred workers.

Bad homophily

Beyond possibly limiting diversity in terms of ethnicity, gender, and other background characteristics, referrals could also serve to perpetuate homophily along other dimensions. One of the main ones to consider is diversity of ideas and opinions. It should be recognized that diversity in terms of background characteristics may be linked to diversity of ideas or may improve other aspects of decision making. For example, one recent study in psychology argues that ethnic diversity increases market efficiency using an experimental asset market. Specifically, they find that market prices are closer to true values when market participants are more diverse.

Homophily may also be undesirable if referrals reflect nepotism. Placing too much value on family and family connections in the hiring (or promotion) process could cause people to lose faith in the idea that their organization is a meritocracy. Examples of nepotism in the hiring process can be found in numerous settings. Using data from Sweden, one study provides evidence that parents often help their children find jobs at the firm where they work [10]. A study using Chinese data shows that men whose fathers-in-law die experience significant drops in earnings, which is consistent with nepotism [11]. The idea is that men receive labor market assistance from their fathers-in-law (including getting help with being considered for higher paying jobs, perhaps in a nepotistic manner) and that such assistance ends upon a death.

It is not always easy to determine whether a source of homophily is “good” or “bad.” Organizations often have distinct cultures for various reasons (such as the founders’ influence, the type of product they provide, or the style and preferences of top managers). Some employees may be an excellent fit in one organization, but may struggle in a different one. Some organizations thrive on a brash, argumentative style, whereas others value polite and measured discourse. While referral hiring may attract individuals who are more like current employees, whether this is bad or good for an organization depends on the individual circumstances at play.