Pepper & Nettle (P&N) provide evidence that lower socioeconomic status (SES) gives rise to a cluster of behaviors – coined “the behavioral constellation of deprivation” (BCD) – that represent contextually appropriate, adaptive responses to the heightened extrinsic mortality and health risks inherent in such conditions. A key feature of P&N's argument is that lower SES engenders limited control over future outcomes. This lack of control reduces the likelihood of realizing delayed rewards, favoring present-focused behaviors.

We appreciate the body of literature demonstrating that inability to control the future contributes to the development and expression of the BCD. However, in much of the research cited here and elsewhere, lack of control is often confounded with unpredictability (e.g., Kidd et al. Reference Kidd, Palmeri and Aslin2013), which is a related but distinct construct (Koolhaas et al. Reference Koolhaas, Bartolomucci, Buwalda, de Boer, Flügge, Korte, Meerlo, Murison, Olivier, Palanza, Richter-Levin, Sgoifo, Steimer, Stiedl, van Dijk, Wöhr and Fuchs2011; Mineka & Hendersen Reference Mineka and Hendersen1985; Nickels et al. Reference Nickels, Cramer and Gural1992). Similar to lack of control, unpredictability is a defining feature of low-SES environments (Mittal & Griskevicius Reference Mittal and Griskevicius2014; Ross & Hill Reference Ross and Hill2002) and plays a key role in shaping physiological responses to environmental stressors (for review, see Koolhaas et al. Reference Koolhaas, Bartolomucci, Buwalda, de Boer, Flügge, Korte, Meerlo, Murison, Olivier, Palanza, Richter-Levin, Sgoifo, Steimer, Stiedl, van Dijk, Wöhr and Fuchs2011). However, the two are distinct constructs, and research examining the unique contributions of each in the development and expression of behaviors consistent with the BCD suggests that environmental unpredictability may be the key driver of these effects. We review this literature below.

The first reason we propose that unpredictability – rather than uncontrollability – drives the BCD is because an outcome's predictability influences one's perception of control over it. Unpredictable events are perceived as more uncontrollable (Glass & Singer Reference Glass and Singer1973; Nickels et al. Reference Nickels, Cramer and Gural1992; Wortman Reference Wortman1975). For example, research has found that individuals subjected to unpredictable – but not predictable – noise blasts exhibited a reduction in perceived control over noise onset, in addition to frustration and impairments in cognition (Glass & Singer Reference Glass and Singer1973). Others have found that participants reported having more influence over exposure to predictable – compared to unpredictable – noise cycles, regardless of whether the participant actually had control (Nickels et al. Reference Nickels, Cramer and Gural1992). Unpredictability's capacity to reduce perceived control is further supported by research examining its role in promoting outcomes associated with the BCD, such as the development of an external locus of control, increased risk taking, and impulsivity (Ross & Hill Reference Ross and Hill2002).

In line with the role that unpredictability plays in shaping one's perceptions of control, one's actual ability to control one's environment depends on the predictability of outcomes available therein (Badia et al. Reference Badia, Harsh and Abbott1979; LaDage Reference LaDage2015; Weiss Reference Weiss1971). Indeed, the advantage of being able to behaviorally control an event is lost if one cannot reliably expect when associated rewards or consequences will arrive. For example, Weiss (Reference Weiss1971) has found less gastric ulceration (a key stress maker) in rats that could escape or avoid shock compared to those that could not, exhibiting the benefit of control. However, even within the group of rats that could exercise control, higher levels of ulceration were found when the onset of the shock was unpredictable than when it was preceded by a signal. Others have found that escape and avoidance behaviors (i.e., attempts to control exposure) were reduced when onset of an aversive stimulus was unpredictable compared to stimuli preceded by a signal (Badia et al. Reference Badia, Harsh and Abbott1979; Dess et al. Reference Dess, Linwick, Patterson, Overmier and Levine1983; Galhardo et al. Reference Galhardo, Vital and Oliveira2011). This appears to result from a learned irrelevance between one's behavioral attempts to control an event and its onset, perhaps analogous to unpredictability's impact on the perception of control in humans (Glass & Singer Reference Glass and Singer1973; Nickels et al. Reference Nickels, Cramer and Gural1992; Overmier & Wielkiewicz, Reference Overmier and Wielkiewicz1983; Wortman Reference Wortman1975). Together, this research suggests that unpredictable events cannot truly be controlled. Therefore, many of the benefits associated with control over future outcomes are erased when those outcomes are temporally unpredictable.

Last, it is important to note that outcome predictability is capable of attenuating the harmful effects of limited control. Being able to predict stressful events decreases how stressful they are, even when one has no control over them (Badia et al. Reference Badia, Harsh and Abbott1979; Dess et al. Reference Dess, Linwick, Patterson, Overmier and Levine1983; Grillon et al. Reference Grillon, Baas, Cornwell and Johnson2006; Martf & Armario Reference Martf and Armario1997). For example, Martf and Armario (Reference Martf and Armario1997) have found that the pituitary-adrenal response of rats became habituated to predictable – but not unpredictable – noise. This suggests that predictability may blunt the stressfulness of unpleasant situations by promoting habituation. Further, predictability can reduce the stressfulness of aversive stimuli by signaling safety, also known as the “safety signal hypothesis” (Seligman & Binik Reference Seligman, Binik, Davis and Hurwitz1977). For example, when a shock arrives predictably, the end of exposure indicates a period of safety that allows for recovery from the previous trial and preparation for the next (Mineka & Hendersen Reference Mineka and Hendersen1985; Seligman & Binik Reference Seligman, Binik, Davis and Hurwitz1977). However, when the stimulus is administered unpredictably, safety is never signaled, enhancing anxiety and hypervigilance (Dess et al. Reference Dess, Linwick, Patterson, Overmier and Levine1983; Herry et al. Reference Herry, Bach, Esposito, Di Salle, Perrig, Scheffler, Lüthi and Seifritz2007; Schmitz et al. Reference Schmitz, Merikangas, Swendsen, Cui, Heaton and Grillon2011; Wieser et al. Reference Wieser, Reicherts, Juravle and von Leupoldt2016). This pattern is also observed in the context of food delivery (e.g., Fokidis et al. Reference Fokidis, des Roziers, Sparr, Rogowski, Sweazea and Deviche2012; Waitt & Buchanan-Smith Reference Waitt and Buchanan-Smith2001). Research in birds has found that when similar amounts of food were presented either predictably or unpredictably, unpredictable delivery was associated with reduced body mass and an increase in circulating glucocorticoids compared to predictable delivery (Fokidis et al. Reference Fokidis, des Roziers, Sparr, Rogowski, Sweazea and Deviche2012).

In summary, we agree that control is a crucial element in determining how individuals respond to the risks in their environments. However, we argue that the ability to predict the future may be a more fundamental factor in facilitating the BCD than control, per se. We propose that considering the contribution of unpredictability to these outcomes could further add to the predictive power of P&N's model.