Worker (dis)trust in management and the delegation of real

Worker (dis)trust in management and the delegation of real
authority∗
Kieron J. Meagher†
Australian National University
Andrew Wait‡
University of Sydney
April 3, 2018
Abstract
Using a unique employee-establishment matched survey, we find a causal relationship between
an individual employee’s distrust of management and the delegation of real authority. We utilize
both fixed effects and instrumental variables to control for unobserved factors: establishmentlevel fixed effects control for management quality, practices, culture and other characteristics;
our instrument of inherited distrust in management addresses the possible endogeneity of employee distrust. Across all specifications, we find that delegation of real authority is less likely
if an employee distrusts management. Our results are consistent with the theoretical literature
on delegation; in particular, when contracts are incomplete distrust between a principal and an
agent prevents delegation that could have been sustained as part of an implicit agreement.
Key words: trust, real authority, delegation, relational contracts.
JEL classifications: D23, L22, L23.
1 Introduction
Firms exist in large part to replace market transactions with (bureaucratic) authority relations
with employees (Williamson, 1985; Grossman and Hart, 1986). The resulting scale, specialization
and division of labour inside modern firms, however, invariably means that valuable information is
distributed across many agents at different levels in an organization’s hierarchy. Furthermore, often
these agents are unable – or unwilling – to fully share this information in a cogent and timely way
(Stinchcombe, 1990; Radner, 1993; Aghion and Tirole, 1997; Jensen and Meckling, 1998; Prendergast,
2002; Dessein, 2002; Meagher et al., 2004). Delegation of decision-making rights is one way to make
use of agents’ information and know-how. As a result, managers do not just supervise employees;
a key task of a manager is to exercise oversight of the delegation of real authority in the face of
non-contractible actions and information.1
∗We would like to thank the following for their useful comments: John Asker, Murali Agastya, Nicholas Bloom,
Federico Ciliberto, Nicolas de Roos, Wouter Dessein, Guido Friebel, Robert Gibbons, Hajime Katayama, Francine Lafontaine, Bentley MacLeod, Kristina McElheran, Michael Raith, Guillaume Roger, Raffaella Sadun, Stefanie Schurer,
Suraj Prasad, Vladimir Smirnov, John Van Reenen and participants at Harvard Business School’s 10 Years of the
World Management Survey conference
†Research School of Economics, H.W. Arndt Building, Australian National University, Canberra, ACT 0200,
Australia. email: [email protected]
‡School of Economics, University of Sydney, NSW 2006 Australia. email: [email protected]
1As outlined in Aghion and Tirole (1997), the primary economic issue of authority inside a firm is who makes a
decision (real authority) as opposed to who has legal authority. Empirically, real authority is not about the formal
rules and processes of an organization but the fuzzy and subjective way in which decision making actually occurs.
1
Delegation, however, is not a panacea for a firm’s contracting problems, as agents must be
encouraged to make appropriate decisions on the basis of their private, non-contractible information.
How well firms are able to encourage agents to take profit-enhancing non-contractible actions is at
the heart of contemporary explanations of “persistent performance differences between seemingly
similar firms” (Gibbons and Henderson, 2013), because an inability to delegate limits a firm’s value
and its potential to grow (Boedker et al., 2011; Bloom et al., 2012).2 Similar ideas are common in
the popular business literature, see for example Williams (2013) and Shockley-Zalabak et al. (2010,
p. 22-23).
One solution to non-contractability could be a relationship in which each party trusts the other
to do the “right” thing.3
In a trusting relationship employees will not (excessively) exploit their
individual authority for their private gain and, in return, management will reward employees appropriately even if they are not legally obliged to do so.
One solution to non-contractability could be a relationship in which the first party trusts the
second party to honour a non-contractible agreement. Specifically, a trust based employment relationship could be one in which employees will take appropriate actions (rather than exploit their
individual authority for their private gain) and, in return, management will reward employees appropriately even if they are not legally obliged to do so.4
The Oxford English Dictionary defines trust as a “firm belief in the reliability, truth, or ability
of someone or something”. This highlights the fact that if a person is to fulfill an obligation they
must have both the intention (honesty) and the ability to do so.
A key contribution of this paper is to conduct the first individual-level empirical analysis of
trust (or its absence) and the delegation of real authority. We use uniquely detailed and matched
employee-establishment data to control for both individual and establishment effects on delegation.
One of our novel results uses an instrumental-variable approach to demonstrate a causal link between
employee (dis)trust in management and the delegation of real decision-making authority.
5
Trust might arise from a number of sources. Culture and social norms affect an individual’s trust
in others. At the personal level, the social capital of the individuals involved in the relationship has
also been shown to contribute to the level of trust (Glaeser et al., 2000). The tacit cooperative
equilibria in the relational-contracting models of Baker et al. (1999, 2002), Levin (2003) and Li et al.
(2017) have been interpreted as trust in a repeated relationship (see Gibbons and Henderson (2013)
for example).6 More behavioral aspects could also play a role, such as the reference points of Hart
and Moore (2008). While theory is framed in the positive – the existence of trust – our empirical
analysis describes the relationship between management and an employee in terms of the employee’s
2Gibbons and Henderson (2013) argue that one driver of performance differences is that better performers manage
to develop organizations in which relational contracts flourish. This, in turn, creates a more efficient organization and
higher productivity. Consistent with this, Boedker et al. (2011) find that of 32 management practices, a measure of
delegation has the highest correlation with their High Performing Workplace Index. Bloom et al. (2012) suggest that
failure to delegate restricts a firm’s ability to grow; they find supporting empirical evidence in a range of manufacturing
firms across a number of countries.
3See the survey by Gibbons (1998) or Malcomson (2013) for a discussion in the context of between-firm behavior.
4The converse could also be a trust based relationship: payment is fixed but the employer trusts the employee to
honour an agreement to do the right thing.
5The theoretical motivations underlying delegation include: a trade-off between costly communication and a loss
of control delegating to a biased agent (Dessein, 2002; Alonso et al., 2008); more effective processing of information by
using agents throughout the organization (Radner, 1993; Bolton and Dewatripont, 1994; Van Zandt, 1999; Garicano,
2000; Meagher, 2003; Rossi-Hansberg and Garicano, 2012); and the incentive effects of providing agents decisionmaking powers (Aghion and Tirole, 1997; Zabojnik, 2002; Bester, 2009; Acemoglu et al., 2007).
6See also the trust game of Kreps (1990).
2
distrust in management – that is the absence of trust. While conceptually one naturally excludes
the other, at the level of survey design distrust is, as we discuss below, a better measure.
More broadly, Arrow (1972) views the issue of trust as pervasive in economic transactions and
its absence as a major cause of poor economic performance. It is now well established empirically
that general trust in society7
is a keystone of economic development, institutional performance
and growth (Greif, 1993; Knack and Keefer, 1997; Algan and Cahuc, 2010; Tabellini, 2010). With
regard to firms, higher national/regional levels of general trust in a society are associated with
larger corporations (La Porta et al., 1997) and more delegation across plants (Bloom et al., 2012).
Gulati and Nickerson (2008) find that preexisting trust between organizations leads to less formal
governance structures and to better performance outcomes in a buyer-seller relationship.8
Micro-founded work on trust and firms is quite rare due to the difficulty of obtaining data. In
a recent study, Macchiavello and Morjaria (2015) find that an ongoing relationship between firms is
important for understanding why supply shocks in the Kenyan rose market are not always exploited.
Within firms, Guiso et al. (2015) and Brown et al. (2015) find a positive relationship between a firm’s
financial performance and the average trustworthiness of its managers as assessed by employees, but
they are unable to identify a mechanism for how trust impacts performance.
Similarly, to date, the empirical literature on decision making has largely sidestepped analyzing
the impact of trust, due mainly to the absence of individual-level data on authority in general, and
trust in particular. The focus has instead been on the business environment. Empirically, Colombo
and Delmastro (2004) find that decisions relating to labor (as compared with capital decisions) are
more likely to be delegated. Acemoglu et al. (2007) find that decentralization is more likely the closer
a firm is to the technology frontier. Bloom et al. (2012) investigate delegation from head office to a
plant using regional measures of trust in general society and societal trust of foreigners. McElheran
(2014) examines the trade-off between delegation (allowing for adaption) and centralization (aiding
coordination) of IT purchases within US manufacturing firms. Bloom et al. (2010) find a positive
relationship between competition and delegation.9 Similarly, Meagher and Wait (2014) emphasize the relationship between delegation and external factors, such as product-market uncertainty,
competition and participation in export markets. Several recent papers examine the complementarity between delegation and other internal practices of an organization. Barrenechea-M`endez et al.
(2016) find a positive relationship between autonomy for blue-collar workers and monitoring of their
activities. Flores-Fillol et al. (2017) find a positive relationship between the delegation of strategic
decisions and teamwork. Hong et al. (2017) that the use of performance pay is associated with
the delegation of authority from the principal/owner to managers, but also the centralization of
authority away from non-managerial employees.10
Our focus on trust, and its absence, at the level of the individual employee has several advantages.
First, theoretical models typically focus on individual principal-agent relationships, not aggregate
relationships between a principal and a work group, or between divisions or establishments in an
7Typically measured with something like the general trust in other people or public institutions questions from
the World Values Survey.
8Using European data, Butler et al. (2016) show that individual trust in others – “[g]enerally speaking, would
you say that most people can be trusted, or that you cant be too careful in dealing with people? – has an inverted
U-shaped relationship to economic performance, as measured by net household income.
9Also see Aghion et al. (2017) for a study relating certain types of delegation to the plant manager from HQ to
better performance during the Great Recession.
10A closely related literature looks at hierarchical structure and reporting (Rajan and Wulf, 2006; Guadalupe and
Wulf, 2010; Guadalupe et al., 2014; Rossi-Hansberg et al., 2014).
3
organization. This means that our empirical study is closely related to the existing theoretical
literature, and in fact we confirm its main prediction: we find a positive relationship between
delegation of real authority to an employee and that employee’s trust of management. Or more
precisely, we find a positive relationship between centralization and distrust of management. Second,
by focusing on individuals, we are able to extend beyond firm-level analysis to examine how outcomes
differ with individual characteristics like tenure, education and gender. Indeed, our results show that
a significant proportion of the observed variation is not explained by establishment (fixed) effects and
that there is sizable and statistically significant within-establishment effect from individual employee
trust.
The instrumental-variable approach we use is based on the following exogenous factor: average distrust in management by the employee’s country of origin. This instrument relates to the
transmission of social norms11 for trust through inheritance. The transmission and persistence of
social norms is itself currently an active area of research. Experiments on the trust game of Berg
et al. (1995) have repeatedly shown participants in the lab follow strategies in one-shot interactions
that cannot have positive payoffs if their opponents maximize monetary payoffs (Camerer, 2003).
Furthermore, the presence of trust measured in these games varies significantly across countries
(Camerer, 2003, p. 87) and on the nationalities of the participants (Glaeser et al., 2000).
Theoretical models, such as Bisin and Verdier (2001), posit the transmission of social norms or
culture both from parents and through contemporaneous external factors such as institutions and
socialization. 12 This gives rise to the persistence of group social norms across time, as observed
empirically by Uslaner (2008), Nunn and Wantchekon (2011) and in the papers surveyed in Bisin and
Verdier (2010). One consequence of this persistence is that immigrants will have trust determined
partly by local experiences and partly by inherited trust from their home country. Algan and
Cahuc (2010) use this fact to generate an exogenous measure of general trust in society; we follow a
similar method to generate our instrument. To the best of our knowledge, our results are the first
examination of incorporating inherited characteristics in a study of employee trust of their workplace
managers. In doing so, we identify a mechanism by which trust affects individual-level economic
activities.
Some of the key results in the paper are as follows. First, from the data is it evident that
there is considerable variation between employees in the real authority they enjoy, both within an
establishment and across the economy. Second, distrust in management also varies substantially
between individuals, even within the same establishment. Taken together, this suggests that both
real decision-making authority and trust relate to the individual, rather than to the establishment
or the firm, validating our approach. Third, as noted above, we find that employee distrust in
management is positively and significantly associated with centralization of real decision-making
authority. Moreover, the magnitude of this relationship is economically important. Finally, this
result continues to hold when we account for: (i) establishment-level fixed effects; and (ii) potential
endogeneity using an instrumental-variable approach.
11We use the term “social norms” broadly to cover the closely related ideas of culture and social capital also
discussed in the literature.
12Similarly, Guiso et al. (2008) use an overlapping-generations model to suggest that, while they update from their
real-world experience, children strongly rely on their parents in formulating their priors regarding the trustworthiness
of others. They also present evidence consistent with their inter-generational transmission prediction. Also see
Dohmen et al. (2012), who find that parental willingness to trust others is correlated with their children’s attitudes
to trust.
4
2 Conceptual framework
In this Section we discuss the key terms of ‘trust’ and ‘delegation’, and outline a simple theoretical
framework that highlights the key empirical question we wish to address.
2.1 What is ‘trust’ ?
As argued by Rousseau et al. (1998), there is no universal definition of trust, but they suggest that
‘[t]rust is a psychological state comprising the intention to accept vulnerability based upon positive
expectations of the intentions or behavior of another’ Rousseau et al. (1998, p. 395).13 Thorelli
(1986, p. 40) defines trust as ‘an assumption on the part of A that if either A or B encounters a
problem in the fulfillment of his implicit or explicit transactional obligations, B may be counted
on to do what A would do if B’s resources would be at A’s disposal’. That is, A trusts B if A
expects B to carry out what they have promised they would do, even in the absence of a formal legal
obligation. Critically, trust is unidirectional because it is a belief about the actions of another party.
According to Sako and Helper (1998), trust can emanate from different underlying sources, such as:
confidence that a party will carry out a formal contractual agreement; that a party is sufficiently
competent that they will be able to complete what was promised; and goodwill that the other party
be willing to undertake mutually beneficial activities and not to take unfair advantage. In many
ways, goodwill could reflect a situation when both parties adhere to a relational contract, as in the
trust game of Kreps (1990), in which parties ‘trust’ one another to fulfill their implicit obligations
given their ongoing relationship and the potential for sufficiently onerous future sanctions. Others
consider trust (or being trustworthy) related to a characteristic, or personality type.
Further diversity of the definition of trust arises in the empirical research. In their survey article,
McEvily and Tortoriello (2011) find 129 different empirical measures of trust (from 179 papers) this
should not be surprising from a game theoretic perspective because beliefs about the actions of other
players/parties are naturally context specific. Our paper is about a particular type of trust; the trust
that a worker has in the statements made by management. Without an enforceable contract it is
this trust – or its absence (distrust) – that will influence the choices a worker makes in response
to a promise (or a threat) by management. As a consequence, trust determines the credibility of
implicit incentives and punishments and henceforth the attractiveness and feasibility of delegating
authority. This worker trust of management relies on several key elements: (i) clarity – the parties
involved must be able to clearly communicate (and understand) what each party is promising to
undertake; and (ii) commitment – if an implicit agreement is to be believed, any promises made
must be credible. This does not necessarily require mutual respect (sometimes emphasized in the
business literature) or even that the transacting parties like one another14; it merely requires that
one party believes that the other will do what they say they will do.15
13In the context of an organization, Mayer et al. (1995) emphasize that there is a trusting party (a trustor) and
a party to be trusted (trustee). The adopt a very similar definition that trust is ‘the willingness of a party to be
vulnerable to the actions of another party based on the expectation that the other will perform a particular action
important to the trustor, irrespective of the ability to monitor or control that other part’.
14Although of course, at the psychological level it is harder for us to trust when our experiences are negative. This
effects goes beyond the obvious Bayesian conclusions about a person’s character from their actions. The ‘halo’ effect
means that we tend to want our views of people to be consistent in all dimensions even if there is no logical reason to
connect the dimensions; an example of this would be thinking of physically attractive people as more talented.
15As the Cuban Missile Crisis so starkly illustrates, alignment of interests is not necessary in order to trust what
someone says. Trusting another person requires that you believe what they say (and believing that what they say
5
…
(a) General trust by management (b) Individual trust of management
Figure 1: Trust in organizations
As noted above, to frame our empirical study, we sketch a reduced-form relational contracting
framework in Section 2.3 to help capture the key relationship between a worker’s trust and the
increased scope for delegation of decision-making authority. Figure 1 further clarifies the type of
trust we focus on. Management might trust (or distrust) workers, as illustrated on the left-hand
side, in Figure 1a – and this will not doubt be important as to whether a principal will want to
delegate to an agent in her organization.16 In this way, Bloom et al. (2012) investigate the relation
between average regional trust – a proxy for individual trust – and delegation from head office to an
establishment. They find that delegation is more likely when the head quarters is in a region with
higher average levels of trust. They also find delegation is more likely to subsidiaries that are based in
highly trusted countries. Rather than management’s trust of workers, our focus in on worker trust of
management; this key difference is illustrated in the right-hand panel in Figure 1b. Concentrating on
individual level trust has several advantages. First, theoretical models typically focus on individual
principal-agent relationships, not aggregated relationships between a principal and a work group,
or between divisions or establishments in an organization. This means that our empirical study is
closely related to the existing theoretical literature. Second, by focusing on individuals, we are able
to examine how outcomes differ between individuals in a particular establishment, extending the
establishment-level analysis of Bloom et al. (2012). Indeed, our results suggest that both delegated
decision-making rights and (dis)trust of management are individual characteristics.
2.2 Real authority and delegated worker autonomy
As pointed out by Aghion and Tirole (1997), formal organizational structures and legal decisionmaking rights do not necessarily coincide with actual decision-making practices inside organizations.
For a variety of reasons, very often the person actually making a decision in a firm is not the one
with the formal (legal) right to decide; it is this real authority that we are interested in, and that
we are able to capture in our data, described below.
As noted in the Introduction, many of the previous studies on decision-making rights have focused
they will do); importantly, it does not require that you to like either message or the messenger. If trust exists, people’s
actions can be directed to more positive equilibria, even in the absence of formal commitments.
16From a revealed preference perspective, delegation to an agent reveals at least some minimum level of trust in
the agent by the principal.
6
on profit-versus-cost centres (Acemoglu et al., 2007) or the delegation of significant decisions between
a higher levels of management and an establishment (Colombo and Delmastro, 2004; Bloom et al.,
2012; Meagher and Wait, 2014; McElheran, 2014). Here we have a complementary focus – we study
how much delegated real authority workers have relating to their work and the work environment
(detailed below in Section 3). Hong et al. (2017) and Flores-Fillol et al. (2017) have a similar focus,
concentrating on these aspects of individual autonomy and influence. Several points are worth noting
here. Firstly, the choice to delegate these decisions to a worker (or not as the case may be) is a
choice of importance for a firm; moreover, the data reveals that there is considerable variation in the
delegated autonomy/decision-making authority workers have, even within the same establishment.
Secondly, the choice regarding delegated autonomy typically relates to a worker’s informal authority;
worker autonomy is most likely to be supported by informal agreements rather than by formal or
legal contracts. This makes trust and distrust potentially very important when it comes to choices
regarding delegation of real authority to workers relating to issues of autonomy. Thirdly, while these
decisions are different from the questions about major changes or procurement studied in Meagher
and Wait (2014) and McElheran (2014) respectively, for example, these aspects of decision-making
authority/autonomy capture some of the essence of what goes on in firms on a day-to-day basis.
In that way, these decisions that we study are very likely to be important, and a likely to have an
impact on the way an establishment operates and performs.
2.3 Some simple formalism
Here we outline a simple reduced-form relational-contracting framework to clarify the economic issues
relating to an individual’s trust/distrust of management, and its implication for the allocation of
decision-making rights. This simple formalism is not intended as a formal theory to test, but rather
clarifies why worker beliefs about management can be important.
Assume a principal P wishes to delegate action a ∈ A to agent A. The payoffs for each individual
are π
i
, i ∈ {P, A}. We assume the action involves no effort cost but the agent’s private benefit is
b(a). Thus b(a) summarizes the incentive conflict between the principal and the agent.
Given this potential incentive conflict, the principal would like to choose an explicit menu m(a)
that compensates/rewards the agent for taking an appropriate action. We assume that, while the
action and any reward schedule are observable to the two parties, they are non-contractible (as are
the payoffs).17
As a reduced form for a relational contract we assume the agent’s trust in the principal to
honestly tell things as they are is τ ∈ {0, 1}, so that τ = 1 is trust in the principal and τ = 0 is
distrust in the principal. In this case, the agent’s optimization problem becomes
arg max
a∈A
Ï€
A = τ (m(a) + b(a)) + (1 − τ )(0 + b(a)), (1)
whereas the principal’s problem is
max
m(·)
Ï€
p
(a
∗
(m, τ )) − m(a). (2)
17By not contractible we also exclude the various mechanisms that might be employed to circumvent the contracting
problem. Recent experimental and theoretical research indicates that these mechanisms are not robust and tend not
to work in practice. See Aghion and Holden (2011) and Aghion et al. (2012), for example.
7
In this non-contractible setting, the principal is able to influence the agent only if the agent
trusts (and understands) the reward schedule announced by the principal.18
Now extend the framework to think of the agent being better informed than the principal about
the payoffs of the action (as in Prendergast (2002) and Dessein (2002)). In this case, the principal has
to choose between taking the action herself with poor information (centralization) versus delegating
the action to the better informed agent. If the principal can influence the agent’s choice of action
through the menu m, delegation becomes relatively more attractive. Thus, if a worker trusts (and
understands) the payoff relevant statements by management, delegation is relatively more attractive.
That is, other things equal, worker trust of management allows for delegation to be sustained as part
of a relational contract where it would otherwise not be possible. This gives us the following empirical
prediction: distrust of management by an individual worker is associated with a lower probability
of delegation of decision-making rights to that worker (i.e. a higher probability of centralization).
This very simple framework could be generalized following the modern contracting literature
to include effort, risk aversion and most importantly some equilibrium enforcement measure for
the ‘trust’ based on repeated interaction. Presumably a repeated game structure with imperfect
monitoring could lead to a breakdown in trust along the equilibrium path, just as price wars occur
under tacit collusion, as in Green and Porter (1984)
An alternative perspective on trust in an organization might be to consider whether a principal
can trust an agent to take the ‘right’ action. This has been formalized through payment congruence
and has generated a large literature (see Aghion and Tirole (1997), for example). Reinterpreting this
literature in terms of trust yields the key insight that the delegation of real authority will only occur
if the principal trusts the agent (has the incentives) to make the ‘right’ choice. However, the main
point we want to make with this simple framework would remain; ‘trust’ that an implicit agreement
will be adhered to by a principal allows for greater delegation of decision-making rights than would
otherwise be feasible.
3 Data Set and Variables
We use the Australian Workplace Industrial Relations Survey 1995 (AWIRS 95) to investigate the
relationship between the delegation of decision-making rights and employee distrust of management.
AWIRS 95 consists of a sample 2001 establishments with 20 or more employees covering all major
industry groups, with the exception of agriculture, forestry, fishing and defence. It also includes
a sample of 19155 employees, randomly selected from the establishments surveyed.19 This data
was collected by the Australian Government Department of Employment and Workplace Relations
(formerly the Department of Industrial Relations).
This paper utilizes data from the General Management Questionnaire, conducted by personal
interview and completed by the most senior manager at the establishment, and the Employee Questionnaire, administered to a randomly selected sample of employees from the surveyed establish18Mayer et al. (1995) emphasize that there is a trustor and a trustee in a relationship in an organization – here
the worker is the trustor who can (potentially) trust management to fulfill its non-contractible promises. In reality
there could also be a second trust relationship with regard to whether management trusts workers to fulfill their
(non-contractible) obligations under delegation.
19AWIRS was a government run survey so establishments were randomly selected from the Australian Bureau of
Statistics Business Register. 2547 establishments were identified to give the 2001 responses. The survey and the data
are described in detail in Morehead et al. (1997).
8
ments. As noted above in the Introduction and detailed below, AWIRS95 has several compelling
attributes. This data set provides a rich description of workers real authority over a range of decisions, and their trust of management. Consequently, this matched employee-establishment data
allows for the first employee-level analysis of trust and delegation. This level of detail regarding
worker autonomy and trust is not available elsewhere, including in the updated version of AWIRS
itself (the Australian Workplace Relations Study). As a consequence, this data source remains relevant, despite its age, as it allows us here to produce new empirical results and insights. Another
advantage of the AWIRS95 data is that it was a carefully designed and implemented survey to reflect
the structure of the broader economy conducted by a government department; the response rates for
the main survey were 80% (Morehead et al., 1997). Furthermore, like the United States, Australia
has a flexible labor market and relatively little interference from unions in the way firms are managed. This means that the observed allocation of decision making is (more) likely to arise from the
optimizing choices of firms, as opposed to being externally imposed by labor-market institutions.
3.1 Centralization of decision-making: dependent variables
Employees were asked ‘[i]n general, how much influence or input do you have about the following?
The type of work you do; How you do your work; When you start and finish work; The pace at
which you do your job; The way the workplace is managed or organised; Decisions which affect you
at this workplace. For each of the six questions, employees can respond: (1) a lot; (2) some; (3) a
little; or (4) none.20
Each of these questions captures related, but different, aspects of the decentralization of decision
authority. Figure 2 shows the influence employees have over these various aspects of their work. In
general, employees have more influence over how they do their job and the pace of work than the
way the establishment is organized or about establishment decisions that affect them.21
To make use of this information, we generate a variable Centralization, by summing the scores
across all six questions and creating a Z-score with a mean of 0 and a standard deviation of 1; lower
values of this variable indicate greater decentralization or employee influence. The recent study
of Flores-Fillol et al. (2017) combine various aspects of worker autonomy and decision authority
into a Z-score. Bloom et al. (2012) also utilize a Z-score index of decentralization. Hong et al.
(2017) consider the allocation of decision-making rights between principals, senior managers and
non-managerial employees relating to 12 operating tasks with some similarities to those considered
here. Further to this, and as noted above, our measure of real authority regarding worker autonomy
complements the focus on the allocation of decision-making rights between higher management levels
and establishments studied in Colombo and Delmastro (2004), Acemoglu et al. (2007), Bloom et al.
(2012) and McElheran (2014).22
20There was also another possible answer of ‘Don’t know’, but only 0.1 percent of employees gave this response.
21The first four questions have their origin in Hackman and Oldham (1975) and have been extensively used over the
last four decades. The statistical validation of these questions as a single measure (of authority), and their distinctness
from empowerment, is covered in Spreitzer (1995). See Thomas (1990) for a discussion of why delegation of decisions
more generally, questions five and six, are also an important component of what economists call authority (and which
he refers to as ‘locus of causality’).
22While there is some subjectivity in these questions, previous studies have also relied on surveys with subjective
questions. For example, Acemoglu et al. (2007) use subjective answers on establishment’s autonomy regarding investment, and on manager’s autonomy/authority regarding employment decisions. Similarly, the measure of delegation
in Colombo and Delmastro (2004) is based on subjective answers from the establishment manager.
9
0 10 20 30 40 Percentage
1 2 3 4
Influence
Type of Work 0 10 20 30 40 50 Percentage
1 234InfluenceHow to do Work010203040Percentage1234InfluenceStart and Finish0 10 20 30 40 Percentage
1 2 3 4
Influence
Pace
0 10 20 30 40 Percentage
1 234InfluenceWorkplace Organization0102030Percentage1234InfluenceDecisions That Affect YouFigure 2: Delegation of decision authority. Key: 1 – A lot, 2 – Some, 3 – A little, 4 – None
As an alternative measure of our dependent variable, we undertake factor analysis of the six
aspects of decision-making autonomy; the first factor has an Eigen value of 2.986, whereas the
other factors have Eigen values of 0.303 or less. We retain the first factor, denoted as Centralization(Factor), and use it as a dependent variable for centralization of decision making. The pairwise
correlation between Centralization and Centralization(Factor) is .995.
Unlike the previous studies described above, we have sufficient power/sample size to repeat our
analysis for each of the individual questions separately as a robustness check in Section 4.4.
3.2 Explanatory variables
Table 1 provides summary statistics for the main variables of interest, including the dependent variables Centralization and Centralization(Factor). Table 2 outlines the pairwise correlations between
the key variables.
Employee characteristics. Employees were asked whether ‘[m]anagement at this workplace can be
trusted to tell things the way they are’. Using this information, we generate a measure of Distrust
equal to 1 if the employee disagrees with the statement and 0 otherwise. We focus on distrust
because the well established negativity dominance (Baumeister et al., 2001; Rozin and Royzman,
2001) in evaluations means that among symmetric good and bad events the bad events dominate
psychologically. As a result more cognitive attention is paid to negative information and behaviors
(Fiske, 1980; Pratto and John, 1991) and learning from/recall of negative events is stronger (Dreben
et al., 1979; Skowronski and Carlston, 1987).
10
Table 1: Summary statistics of basic sample (N = 16922)
Variable Mean Std Dev.
Dependent variable
Centralization -0.018 0.997
Centralization (Factor) -0.017 0.922
Type of work 2.293 1.052
How do work 1.795 0.929
Start/finish time 2.542 1.208
Pace of work 2.039 1.053
Establishment organization 2.965 1.054
Decisions that affects you 2.698 0.998
Key explanatory variables
Distrust 0.336 0.472
Male 0.563 0.496
Tenure 6.285 6.849
Tenure2
86.401 182.825
Instrument
Inherited distrust 0.333 0.025
Notes: a. Source AWIRS 95.
Table 2: Correlations between key variables (N = 16922)
Variable Centralization Distrust Male Tenure Inherited distrust
Centralization 1.000
Centralization (Factor) 0.995 0.257
Distrust 0.253 1.000
Male -0.032 0.075 1.000
Tenure -0.026 .109 0.160 1.000
Inherited distrust 0.023 0.038 0.019 -0.002 1.000
Type of work 0.794 0.193 -0.027 -0.037 0.022
How do work 0.770 0.151 -0.010 -0.019 -0.013
Start/finish time 0.735 0.145 -0.015 0.001 0.038
Pace of work 0.740 0.159 0.016 -0.006 0.004
Establishment organization 0.770 0.241 -0.074 -0.046 0.041
Decisions that affects you 0.767 0.275 -0.036 -0.017 0.007
Notes: a. Source AWIRS 95.
We create a dummy variable if an employee is Male (1) or otherwise (0). Variables were created indicating each employee’s Tenure and their tenure squared (Tenure2
) at the establishment.
Similarly, a series of dummies are generated for the employee’s age and their highest level of education. Dummy variables for the employee’s occupational group were also constructed. These three
sets of dummy variables are included in most of the estimates below. The pairwise correlations between Centralization and Distrust and the education variables (Table 9) and the occupation dummy
variables (Table 10) are provided in Appendix A.
11
4 Empirical results
The key relationship we examine is between employee distrust of management and the allocation
of decision-making authority. Figure 3 highlights some of the important features of Distrust and of
our dependent variable Centralization. The top left-hand panel in the figure plots the distribution
of Centralization with the Normal distribution overlaid. This, along with Figure 2, shows there is
substantial variation in decision-making authority across employees.
The second panel, in the top right-hand corner, plots the distribution of average establishmentlevel Distrust. Since Distrust is a binary variable, the average of Distrust within a establishment
is the proportion of employees who distrust management. From this, it is evident that within
most establishments there is considerable diversity in employee distrust of management, suggesting
distrust is an individual, not an establishment, characteristic. This highlights a significant advantage
of our approach: we explore (dis)trust and its relationship to delegation at the employee level.
The lower left-hand panel of Figure 3 shows the kernel-density plots of the within-establishment
variation (the dotted line) and the between-establishment variation (the black line) of Centralization.
23 While both densities are centred around 0, the within-establishment distribution has a
greater variance than the between-establishment distribution. This suggests there are not simply
decentralized or centralized establishments in the economy. Rather, there are important differences
in decision-making authority/autonomy within establishments.24
The final panel in the figure, in the bottom right-hand corner, shows the relationship between
Centralization and employee Distrust of management. It shows the kernel-density of Centralization
for employees who distrust management (indicated by the blue line) and for employees who do not
distrust management (the red line). This show there is less delegation to employees who distrust
management. We analyze this key relationship further below.
4.1 Distrust and delegation within establishments: Fixed-effects estimation
To explore the relationship between an employee’s distrust of management and delegation, we estimate both pooled OLS and establishment fixed-effects (FE) models. The OLS estimating equation
is
ci = β1di + βXi + εi
, (3)
where ci
is the centralization index, di
is Distrust and Xi
is a vector of individual characteristics and controls as described previously. The fixed-effects specification adds a constant uj[i]
for
each establishment j, where the function j[i] identifies the establishment j of which person i is an
employee:25
ci = β1di + βXi + uj[i] + εi
. (4)
Table 3 displays the estimated coefficients for Distrust and the other key explanatory variables
of interest. While not implying causality, these results provide insight into the relationship between
23The within-establishment densities are calculated after subtracting the establishment mean. The betweenestablishment density compares establishment means.
24A range of levels of autonomy/empowerment among similar workers doing similar jobs in the same organization
has been observed elsewhere, notably by Spreitzer (1995). Also see Hackman and Oldham (1975).
25This is equivalent, but more concise, than the double-index notation: cij = β1dij + βXij + uj + εij
12
0 .1 .2 .3 .4 Density
-2 -1 0 1 2Centralization
0.511.522.5Density0.2.4.6.81Average plant distrust0 .2 .4 .6 .8 Density
-4 -2 0 2 4Centralization
Within Between0.1.2.3.4Density-2-1012CentralizationDistrust No distrustFigure 3: Centralization of decision authority and employee trust
distrust and delegation.
Model (1) shows the estimated OLS coefficients for our key explanatory variables, using Centralization as a dependent variable, without any other controls. Model (2) shows the pooled OLS
estimation results (no establishment-level fixed effects) with the addition of education, employee age
and occupation controls. The results show Distrust associated with less delegation (higher values
of Centralization). This is consistent with the predictions of the relational-contracting literature.
Introducing individual-level characteristics (gender, tenure, education, age and occupation) does not
change the basic insight from Figure 3 – individuals who Distrust management have on average fewer
decision-making rights.
The estimated coefficient for Male is negative and significant – males are more likely to be
delegated decision-making rights, other things equal. In addition, the estimated coefficients for
Tenure and Tenure2
show it has a negative, convex and significant relationship with centralization.
Tenure is typically interpreted as a proxy for the accumulation of an employee’s firm-specific human
capital; tenure allows an employee to accumulate the requisite specific information to make effective
decisions.
Individual establishment managers no doubt have their own style; it is possible, for example, that
some managers will have a greater propensity to delegate than others. They will also differ in their
abilities and in the other work practices they adopt.26 Consequently, it is important to account
for omitted establishment characteristics. To do this we estimate establishment-level fixed-effect
26For example, McElheran (2014) finds differences in the delegation of purchasing authority to different establishments, even within the same organization.
13
models, as shown in Models (3) and (4).
The overall R2
reported for Model (3) indicates that the observed characteristics of individual
employees explain approximately 20.7% of the variation in Centralization. The remaining 79.3% of
the variation is a residual attributed to the fixed effects and the idiosyncratic error term. The reported ρ shows that the fixed effects account for approximately 19.1% of the residual 79.3% variation,
that is 15.1% of the total variation. Thus for the within-establishment model the observed employee
variables make a moderately larger contribution, approximately one third larger, in explaining the
variation of Centralization, than the fixed effects.
Overall these results show that both establishment and individual employee characteristics are
important for understanding delegation. Thus it is imperative to extend the existing establishmentlevel models to include information on employees and their relationship with management.
The estimates from Model (3) show a strong and significant positive relationship between Centralization and employee Distrust of management. This is an important empirical finding; controlling
for establishment-level fixed-effects, there is a significant relationship between Distrust and the allocation of decision authority within an establishment. The estimated impact of distrust is large:
almost three-times the magnitude of having an postgraduate degree (compared with not finishing
high school) but only three quarters of the impact of being a professional vis-a-vis a laborer (Table
3, discussed further below).
Furthermore, using establishment-level fixed effects have a modest impact on the estimated
coefficients for Distrust, Male and Tenure both in terms of their magnitude and significance. The
coefficients for Tenure and Tenure2
increase in magnitude upon the inclusion of establishment-level
fixed effects. These results suggest that while managerial ability and establishment practices are
important, the relationship between employee distrust and delegation remains – if an employee trusts
management, there is a higher probability of delegation to that individual.
As a robustness check, Model (4), also in Table 3, uses the alternative dependent variable of the
first factor, Centralization(Factor). These estimates again include the establishment fixed effects.
The results are very similar to the estimated coefficients in the first two models; crucially, the
relationship between Distrust and delegation remains significant.
Table 3 also reports the estimated coefficients for education, occupation and age for Models (2),
(3) and (4). The impact of the establishment level fixed effects on the estimated coefficients for the
education, occupation and age categories can be seen by comparing the results in Models (2) and
(3).
The probability of delegation increases strongly, on average, with highest educational attainment
of the employee (Table 3, Model 2), however this effect is substantially smaller within an establishment (Model 3). This suggests that the differences in delegation across educational levels in the
labor market are in part due to sorting: establishments with more highly educated employees are
more decentralized on average. Including establishment fixed effects also reduces the size of the
estimated coefficients for the age dummy variables (Table 3, Models 3 and 4).
14
Table 3: Centralization of decision-making (standard errors in parentheses)
a
(1) (2) (3) (4)
OLS OLS ESTABLISHMENT FE ESTABLISHMENT FE
Dep. variable Centralization Centralization Centralization Centralization(Factor)
Distrust 0.556*** (0.017) 0.511*** (0.016) 0.484*** (0.016) 0.459*** (0.015)
Male -0.089*** (0.017) -0.065*** (0.017) -0.047*** (0.018) -0.045*** (0.017)
Tenure -0.013*** (0.003) -0.011*** (0.003) -0.016*** (0.003) -0.015*** (0.003)
Tenure
2/1000 0.256** (0.107) 0.175* (0.099) 0.292*** (0.105) 0.274*** (0.099)
Education
Year 10 -0.105* (0.055) -0.103* (0.058) -0.101* (0.054)
High School -0.152*** (0.058) -0.122** (0.060) -0.116** (0.056)
Basic vocational -0.127** (0.065) -0.105 (0.068) -0.098 (0.063)
Skilled vocational -0.135** (0.059) -0.123** (0.062) -0.124** (0.057)
Assoc. diploma -0.219*** (0.060) -0.144** (0.063) -0.135** (0.059)
Undergraduate -0.179*** (0.060) -0.126** (0.064) -0.118** (0.059)
Postgraduate -0.206*** (0.063) -0.168** (0.066) -0.156** (0.061)
Other -0.143* (0.076) -0.123 (0.078) -0.125* (0.073)
Age (Years)
21 – 24 -0.093** (0.038) 0.002 (0.041) -0.007 (0.038)
25 – 29 -0.198*** (0.037) -0.090** (0.040) -0.092** (0.037)
30 – 34 -0.225*** (0.038) -0.100** (0.042) -0.094** (0.039)
35 – 39 -0.203*** (0.038) -0.087** (0.042) -0.081** (0.039)
40 – 44 -0.192*** (0.039) -0.076* (0.042) -0.071* (0.039)
45 – 49 -0.194*** (0.039) -0.090** (0.043) -0.083** (0.040)
50 – 54 -0.142*** (0.043) -0.035 (0.048) -0.031 (0.044)
55+ -0.168*** (0.047) -0.093* (0.051) -0.078* (0.047)
Occupation
Machine operators -0.051 (0.036) 0.009 (0.037) -0.006 (0.035)
Sales & personal service -0.173*** (0.033) -0.213*** (0.037) -0.198*** (0.034)
Clerks -0.405*** (0.030) -0.342*** (0.034) -0.306*** (0.032)
Tradespersons -0.225*** (0.036) -0.209*** (0.039) -0.211*** (0.035)
Para-professionals -0.354*** (0.035) -0.420*** (0.038) -0.396*** (0.035)
Professionals -0.639*** (0.036) -0.627*** (0.039) -0.580*** (0.036)
Managers -1.269*** (0.032) -1.197*** (0.037) -1.112*** (0.034)
Other -0.454*** (0.098) -0.496*** (0.089) -0.464*** (0.084)
R2 0.069 0.210 0.207 0.206
σui 0.417 0.379
ρ 0.191 0.185
a No. of observation = 16922. Clustered standard errors for 1773 establishments. *** significant at 1% level, ** significant at
5% level, * significant at 10% level. ρ = (σuj[i]
)
2/[(σuj[i]
)
2 + (σεi
)
2] where σεi is the standard deviation of residuals of the
overall error term εi. σuj[i] is the standard deviation of the group residuals uj[i]. For Models (3) and (4) the overall
R2 is
reported, which excludes the effect of the fixed effects. Omitted categories: Highest level of education less than Year 10; Age
between 15 and 20; and Occupational group Laborer (and related).
15
As one might expect, occupations such as manager and professional have, ceteris paribus, the
strongest link with delegation. Furthermore, the relationships between occupations and delegation
are largely unaffected by establishment fixed effects.
While the motives for delegation are not observed in our data, these results are consistent with
the theory that any advantages from decentralization can be more readily realized when there is
trust between an employee and management. Of course, while our estimates are broadly consistent
with these theoretical predictions, our results are not a full test of any specific theory; in particular,
these empirical results do not describe the underlying conditions that help facilitate a ‘trust’-based
equilibrium. Furthermore, we are yet to account for the potential endogeneity between delegation
and distrust – we turn to this issue now.
4.2 Instrumental-variable approach
Our estimates in the previous section show a statistically significant and economically important
relationship between employee distrust and delegation. This relationship remains after controlling for
establishment-level fixed effects that could include factors such as managerial quality, management
practices or culture. However, there is still the potential for the endogeneity of employee’s distrust.
As is typical, three types of endogeneity are of potential concern: measurement error; omittedvariable bias; and simultaneity. Below we follow the standard approach (Angrist and Pischke, 2008)
and use an instrumental-variables (IV) approach to address endogeneity. The possible omitted
variable bias and the simultaneity of Distrust and Centralization follow well documented patterns
and can be resolved by finding appropriate instruments (Gujarati, 2005; Wooldridge, 2000). The
possibility of measurement error, which are more closely related to the details of our data, merits
further discussion.
Measurement error could impact our estimated results in several ways. First, there could be noise
in the reporting of the questions making up Centralization, our decision-making allocation measure.
This noise will be mitigated to some degree by averaging the six questions.27 The impact of the
residual noise in the dependent variable is to decrease the goodness of fit in our estimation, biasing
us against finding statistically significant results. Second, any noise in the reporting of Distrust
would produce attenuation, biasing the coefficient on Distrust towards zero. Thus in both of these
cases the potential bias works against finding a significant relationship between (de)centralization
and trust.
A measurement-error issue that could arise from our subjective variables is common-methods
or survey bias. Survey bias could occur if there exists a common factor si that distorts individual
responses to the key subjective questions in the survey. This might be due to the survey itself (an
individual is more positive in answering questions face-to-face) or due to idiosyncratic factors like
feeling tired/sick or individuals may exhibit the psychology trait of making their responses unrealistically consistent. This last effect is often referred to as the ‘halo effect’ because positive assessments
of one characteristic typically spill over into positive assessments of unrelated characteristics (Kahneman, 2011).
We develop the following model, based on classical errors-in-variables (CEV) framework, as in
Wooldridge (2000), but extended to a common factor for the error, in order to explore our survey
27See Section 4.4 below for a robustness check relating to the construction of our decision-authority variable.
16
bias issue. Specifically assume that both ‘true’ distrust, d
∗
i
, and true ‘centralization’ c
∗
i
are mismeasured due to a common factor (mean zero) survey effect si so that we actually observed di and
ci
:
di = d
∗
i + α1si and ci = c
∗
i + α2si
. (5)
The standard CEV assumption gives corr(d
∗
i
, si) = 0 and corr(c
∗
i
, si) = 0, and the true relationship
(omitting establishment fixed effects for convenience) is
c
∗
i = β1d
∗
i + βXi + i
, (6)
where X is the vector of variables unaffected by measurement error. Except for the measurement
error we assume that the standard regression assumptions are satisfied: in particular is uncorrelated
with c
∗
, d∗
, X and s. Substituting for the observed variables gives
ci − α2si = β1(di − α1si) + βXi + i
. (7)
Rearranging we get
ci = β1di + βXi + (α2 − β1α1)si + i = β1di + βXi + ei
, (8)
where γ = α2 − β1α1 and ei = γsi + i
. In our case α1 and α2 plausibly have the opposite sign,
and thus the sign of the bias arising from the common factor survey bias would be negative. Thus,
common factor survey bias is a particular form of omitted variable bias.
We use an IV approach to account for endogeneity, including omitted variable bias. Our candidate
instrument is distrust of management by country of birth.
An employee’s country of origin potentially affects their distrust in management due to the
social reproduction of norms as described by Bisin and Verdier (2001). That is, the institutions, and
culture of where someone was born could help shape their trust in others and their overall attitude
to authority – see Bidner and Francois (2011) and Bloom et al. (2012), for example.28 We calculate
the average level of distrust for employees from each country of birth, excluding all observations from
each employee’s own establishment or establishment.29 The variable – Inherited Distrust – is used
as our instrument for an employee’s distrust in management. A key advantage of this instrument
is that potential endogeneity between employee distrust and delegation within an establishment is
ruled out because an employee’s own establishment is excluded by definition. Thus, there are strong
theoretical reasons to believe Inherited Trust is exogenous in the second-stage regression.30
Instrumental Validity and Identification
The IV estimates are reported in Table 4 (first-stage) and Table 5 (second-stage). In both Mod28In a different context, Blau et al. (2013) and Fern´andez and Fogli (2009) use country of origin as a proxy for
gender role attitudes.
29Due to degrees of freedom issues, countries from the same region with small numbers of employees in the sample
are pooled.
30One might be concerned that Inherited Distrust is in someway a proxy for employer racism. This seems unlikely
since groups typically discriminated against under the White Australia policy (Asians, Africans and people from the
Middle East) have lower rates of distrust than those who the race based policies favoured (UK, NZ and USA).
17
Table 4: IV First-Stage results Centralization of decision-making: OLS and establishment fixedeffects estimation coefficients (standard errors in parentheses)a
(5) (6)
OLS-IV ESTABLISHMENT FE-IV
Dep. Variable Distrust Distrust
Inherited distrust 0.837*** (0.144) 0.970*** (0.151)
Male 0.041*** (0.009) 0.025*** (0.009)
Tenure 0.021*** (0.001) 0.017*** (0.001)
Tenure2
/1000 -0.501*** (0.052) -0.425*** (0.053)
Education
Year 10 -0.009 (0.025) -0.007 (0.025)
High School 0.035 (0.025) 0.034 (0.025)
Basic vocational 0.044 (0.030) 0.034 (0.031)
Skilled vocational 0.069** (0.027) 0.059** (0.028)
Assoc. diploma 0.041 (0.028) 0.041 (0.028)
Undergraduate 0.041 (0.027) 0.029 (0.028)
Postgraduate 0.071** (0.029) 0.053* (0.029)
Other 0.036 (0.035) 0.039 (0.035)
Age (Years)
21 – 24 0.098*** (0.017) 0.077*** (0.019)
25 – 29 0.132*** (0.017) 0.100*** (0.019)
30 – 34 0.139*** (0.017) 0.096*** (0.019)
35 – 39 0.122*** (0.018) 0.086*** (0.019)
40 – 44 0.097*** (0.017) 0.057*** (0.019)
45 – 49 0.062*** (0.018) 0.021 (0.020)
50 – 54 0.054*** (0.019) 0.008 (0.021)
55+ -0.020 (0.021) -0.048** (0.022)
Occupation
Machine operators 0.050*** (0.018) 0.038** (0.019)
Sales & personal service -0.060*** (0.015) -0.044*** (0.017)
Clerks -0.010 (0.015) -0.061*** (0.016)
Tradespersons 0.051** (0.020) 0.048** (0.021)
Para-professionals 0.052*** (0.018) 0.004 (0.018)
Professionals -0.036** (0.018) -0.053*** (0.019)
Managers -0.177*** (0.017) -0.182*** (0.018)
Other -0.040 (0.039) -0.045 (0.042)
Kleibergen-Paap F-statisticc
33.888** 41.023**
5% Critical values (r = 10%) 16.38 16.38
No. of obs. 16922 16922
a *** significant at 1% level, ** significant at 5% level, * significant at 10% level. Clustered standard
errors in parentheses for 1773 establishments.
b The second-stage IV estimates for Centralization, the dependent variable in both models, are shown in
Table 5.
c Weak instruments test: Kleibergen-Paap rk Wald F statistic, H0 = weak instruments. Critical values
from (Stock and Yogo, 2005, Table 2).
18
Table 5: Second-stage IV results Centralization of decision-making: OLS and establishment fixedeffects estimation coefficients (standard errors in parentheses)a
(5) (6)
OLS-IV ESTABLISHMENT FE-IV
Dep. Variable Centralization Centralization
Distrust 1.743*** (0.424) 1.323*** (0.361)
Male -0.116*** (0.027) -0.069*** (0.023)
Tenure -0.037*** (0.009) -0.030*** (0.007)
Tenure2
/1000 0.790*** (0.248) 0.647*** (0.196)
Education
Year 10 -0.110* (0.062) -0.107* (0.061)
High School -0.205*** (0.067) -0.157** (0.065)
Basic vocational -0.194** (0.078) -0.142* (0.075)
Skilled vocational -0.232*** (0.075) -0.180** (0.071)
Assoc. diploma -0.277*** (0.071) -0.183*** (0.070)
Undergraduate -0.234*** (0.071) -0.153** (0.070)
Postgraduate -0.296*** (0.078) -0.212*** (0.074)
Other -0.188** (0.091) -0.155* (0.087)
Age (Years)
21 – 24 -0.213*** (0.062) -0.062 (0.053)
25 – 29 -0.355*** (0.067) -0.170*** (0.054)
30 – 34 -0.390*** (0.071) -0.175*** (0.056)
35 – 39 -0.346*** (0.065) -0.154*** (0.054)
40 – 44 -0.304*** (0.057) -0.117** (0.048)
45 – 49 -0.261*** (0.050) -0.101** (0.047)
50 – 54 -0.202*** (0.052) -0.036 (0.051)
55+ -0.135** (0.056) -0.046 (0.059)
Occupation
Machine operators -0.115** (0.047) -0.024 (0.043)
Sales & personal service -0.107** (0.044) -0.180*** (0.041)
Clerks -0.399*** (0.036) -0.294*** (0.041)
Tradespersons -0.294*** (0.048) -0.252*** (0.045)
Para-professionals -0.429*** (0.049) -0.429*** (0.040)
Professionals -0.608*** (0.045) -0.590*** (0.045)
Managers -1.063*** (0.079) -1.054*** (0.072)
Other -0.408*** (0.107) -0.460*** (0.095)
Observations 16922 16922
a Clustered standard errors for 1773 establishments estimated in parentheses. *** significant at 1% level,
** significant at 5% level, * significant at 10 % level. First-stage estimates shown in Table 4.
19
els 5 and 6 the (endogenous) dependent variable in the first stage is Distrust and the dependent
variable in the second stage is Centralization. Model 6 reports the IV results accounting for establishment fixed effects. Statistical tests on the appropriateness of the IV approach are given at the
bottom of Table 4.
The instrument performs well; it is significant in the first-stage regressions for both the IV and
fixed-effects-IV specifications. The coefficient sign accords with the theories presented above: higher
Inherited Distrust is associated with higher individual distrust.
The validity of inference on the Distrust coefficient in the second-stage IV regressions requires
sufficiently ‘strong’ identification (not weak instruments). For clustered standard errors the appropriate first stage F-test for the significance of the instruments is the Kleibergen-Paap rk Wald F
statistic (Kleibergen and Paap, 2006) since it allows for heteroskedasticity in calculating the matrix
rank. The corresponding critical values, given at the bottom of Table 4 are from Stock and Yogo
(2005).31 For both models, (5) and (6), the hypothesis of weak identification/weak instruments is
strongly rejected at the 5% level of significance with an r of 10%, implying that the standard causal
inference on the coefficient for Distrust should be valid.
Estimation Results
The first-stage results for the other controls in Table 4 are economically interesting in their own
right. There are gender differences in distrust (males distrust more) and while these are reduced
by about 40% when we include establishment fixed effects they are still significant. Tenure has
a concave relationship with distrust and is positive and increasing for most employees.32 Thus it
appears that, on average, employees’ trust of management is eroded rather than built over time.33
While distrust increases with tenure it decreases strongly with age between 21 years and retirement. Since this is a cross sectional data set we cannot separate the effects and aging from birth
cohort effects.
The second-stage IV estimates indicate a negative and highly significant relationship between
Distrust and delegation. The estimated coefficient increases in absolute size compared with the
estimated Distrust coefficients in Table 5. The Distrust coefficient is larger than the effects for most
of the other employee characteristics but it is a similar magnitude to some of the inter occupational
differences in Distrust.
The change in the Trust coefficients between Table 3 and 5 indicates the importance of addressing
endogeneity. Of course, we do not know the specific form of the endogeneity, however, as discussed
above one potential candidate is behavioral: a ‘halo’ effect could produce common factor survey
bias. Nevertheless, across all specifications, distrust is significantly negatively related to delegation.
The key question from the various perspectives on trust and delegation discussed previously,
such as the the relational-contracting literature, is whether the Distrust coefficient is positive. The
31The critical value is a function of the number of included endogenous regressors (n = 1), the number of instrumental variables (K2 = 1), and the desired maximal size (r) of a 5% Wald test of β = β0. An r of 10% is the strictest
level of r reported in Stock and Yogo (2005). The test has a null hypothesis of weak instruments, see Kleibergen and
Schaffer (2007) regarding implementation.
32In Model (6), for example, the turning point of the quadratic for tenure is around 20 years. And about 5% of the
estimation sample have more the 20 years of tenure.
33It is tempting to try and build a better instrument by interacting inherited distrust with tenure to capture the
fact that the impact of exogenous inherited distrust is weakest when individuals’ initially join a firm and subsequently
evolves over time as the first stage estimates indicate. The problem with this approach is that as the second stage
estimates indicated time/tenure also has a significant correlation with centralization so the exclusion restriction would
fail to hold for an instrument made by the interaction.
20
instrumental variable approach has a number of well known issues, including the increase in the
standard error around the estimate of the endogenous coefficient (Wooldridge, 2000). In addition,
taking the heterogenous treatment effects perspective (Stock and Watson, 2015) the IV estimate is
the Local Average Treatment Effect (LATE) rather than the Average Treatment Effect (ATE). The
LATE estimate of the Distrust coefficient is a weighted average of the individual effect of distrust on
delegation, with more weight attached to those most affected by the instrument Inherited Distrust,
that is for who past experiences most impact their current (dis)trust. The determinants of trust in
economic contexts is an ongoing area of research but this could plausibly lead to a LATE estimate
greater than the ATE estimate.
In all, however, the critical question we address in the paper is not the point estimate of Disrust,
but rather the probability that the estimated coefficient is positive and bounded away from zero.
Indeed, we find the estimated coefficient is statistically significantly positive in all of our specifications, including our pooled and fixed-effects estimates, IV models and even when we use alternative
ordinal measures of centralization (below in Section 4.4).
4.3 Robustness of results
As shown above, across a range of specifications we find a significant and economically important
negative relationship between distrust and the delegation of decision-making authority to individual
employees. Here we present two more robustness checks: (i) an examination of the relationship
between trust and delegation for non-managers only; (ii) re-examining the relationship between
trust and decision authority in the services sector only; and (iii) estimating the trust/delegation
relationship for each of our six measures of autonomy separately.
4.3.1 Real authority and trust for non-managers and in the services sector
The estimates above include all occupational groups surveyed, including managers. While these
managers surveyed in the employee questionnaire of AWIRS are lower-level managers – they are
not the general manager of the establishment or head of employment relations for example – as
a robustness check we re-estimate our preferred establishment fixed-effects instrumental-variables
model excluding all employees who nominated themselves as being part of the manager occupation
group.34 As shown in Model (7), Table 6, our key results hold; there remains a strong negative
relationship between worker distrust and delegation.
AWIRS is a cross-industry survey that includes establishments from mining, manufacturing and
the services sector. While we have included fixed effects to control for establishment effects, no
doubt technical requirements will differ across different industries; for example, technical reasons
might dictate that someone working in manufacturing has less autonomy over the pace of work or
the time they start than an employee in some roles in the services sector. To account for these
possible issues, we re-estimate Model (6) for only the services-sector establishments (excluding any
establishment from the mining and manufacturing sectors). These results – Model (8) in Table 6 – are
very similar to the estimates obtained using the full sample, reaffirming the statistically significant
and economically important relationship between distrust and decision-making allocation.
34As Table 10 shows managers have higher distrust than every other occupation except for professionals.
21
Table 6: Decision making and Distrust for subsamples, selected IV estimates with establishment
fixed effects (Clustered standard error in parentheses)a
(7) (8)
Subsample Non-managers onlyb Service sector onlyc
Coeff (SE) Coeff (SE)
First-stage Estimates (Distrust)
Instrument
Inherited Distrust 0.945∗∗∗ (.161) 0.865∗∗∗ (.183)
Weak Instrument Test
Kleibergen-Paap F-statisticd
34.26∗∗ 22.44∗∗
5% Critical values (r = 10%) 16.38 16.38
Second-stage estimates (Centralization)
Distrust 1.555∗∗∗ (.416) 1.103∗∗ (.466)
No. of obs. 15415 12930
a *** Significant at 1% level, ** significant at 5% level, * significant at 10 % level. Models (7) and
(8) are equivalent to fixed-effects IV Model (6) with a full set of education, age and occupational
controls.
b Excluding individuals whose occupation is manager. Clustered standard errors for 1762 establishments shown in parentheses.
c Excluding establishments in the mining and manufacturing industries. Clustered standard errors
for 1437 establishments shown in parentheses.
d Weak instruments test: Kleibergen-Paap rk Wald F statistic, H0 = weak instruments. Critical
values from (Stock and Yogo, 2005, Table 2).
4.4 Robustness of decision-authority measure: an ordinal approach
As described above in Section 3, our main dependent variable is a composite of six separate aspects
of an employee’s decision-making authority: type of work; how to do the work; start and finish
times; pace of work; influence regarding workplace organization; and influence on decisions in the
firm that affect the employee. Here, we explore how the estimated relationship between distrust and
decision-making authority is driven by each one of these six measures separately.
Each of these six measures is scored on a four point scale measuring the amount of influence the
employee has over that type of decision/activity: (1) A lot, (2) Some, (3) A little, and (4) None.
Rather than impose a linear scale in our modeling we instead take an ordinal approach and estimate
the second-stage of our instrumental-variables model as an ordered probit. Thus in the following we
relax both the pooling and linearity assumptions that were employed to construct the Z-scores.
While the ordered probit allows us to relax the cardinality of Centralization, as it is a nonlinear model it precludes the use of the within estimator to calculate the establishment-level fixed
effects. Thus the following results reported in Table 7 are corrected for the endogenity of Distrust by
standard instrumental-variable methods (IV) but not for establishment fixed effects. As in Models
(5) and (6), this estimation includes all age, education and occupation controls, the instrument
Inherited Distrust and clustered standard errors. For the sake of brevity, only the key estimated
coefficients are reported here.
The results in Table 7 suggest that the negative relationship between Distrust and delegation
22
Table 7: Centralization of decision-making: IV Ordered Probit model of individual variables coefficients (standard errors in parentheses)a
(9) (10) (11) (12) (13) (14)
Second Stage Ordered Probit
Dep. Variable Type How do Start/finish Pace of Establishment Decisions that
of work work times work Organization affect you
Male -0.082*** 0.018 -0.160*** -0.006 -0.155*** -0.045*
(0.022) (0.028) (0.022) (0.027) (0.020) (0.027)
Tenure -0.042*** -0.009 -0.031*** -0.015* -0.040*** -0.011
(0.005) (0.010) (0.006) (0.009) (0.005) (0.009)
Tenure2
/1000 0.981*** 0.193 0.720*** 0.272 0.900*** 0.101
(0.164) (0.257) (0.168) (0.235) (0.149) (0.248)
Distrust 1.598*** -0.046 1.751*** 0.748* 1.973*** 0.911**
(0.260) (0.437) (0.182) (0.389) (0.147) (0.411)
First Stage IV (Distrust)
Inherited distrust 0.837*** 0.837*** 0.837*** 0.837*** 0.837*** 0.837***
(0.144) (0.144) (0.144) (0.144) (0.144) (0.144)
Male 0.041*** 0.041*** 0.041*** 0.041*** 0.041*** 0.041***
(0.009) (0.009) (0.009) (0.009) (0.009) (0.009)
Tenure 0.021*** 0.021*** 0.021*** 0.021*** 0.021*** 0.021***
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
Tenure2
/1000 -0.501*** -0.501*** -0.501*** -0.501*** -0.501*** -0.501***
(0.052) (0.052) (0.052) (0.052) (0.052) (0.052)
Cut 1 -0.956*** -0.893*** -0.728*** -0.674*** -1.169*** -1.431***
(0.174) (0.079) (0.158) (0.107) (0.209) (0.115)
Cut 2 -0.109 0.015 -0.223** 0.146 -0.536*** -0.391***
(0.096) (0.095) (0.102) (0.089) (0.135) (0.099)
Cut 3 0.418*** 0.714*** 0.090 0.716*** -0.053 0.498***
(0.073) (0.113) (0.078) (0.081) (0.089) (0.089)
No. of obs. 16922 16922 16922 16922 16922 16922
a
*** significant at 1% level, ** significant at 5% level, * significant at 10 % level. Clustered standard errors
in parentheses for 1773 establishments. All models include controls for education, age and occupation.
23
Table 8: Average marginal effects of one unit change in Distrust on individual centralization
of decision-making variables: Ordered Probit model with IV of Distrust. (standard errors in
parentheses)a
(9) (10) (11) (12) (13) (14)
Dep. Variable Type How do Start/finish Pace of Establishment Decisions that
of work work times work Organization affect you
Influence
1. A lot -0.429*** 0.017 -0.485*** -0.264** -0.376*** -0.148**
(0.072) (0.162) (0.055) (0.128) (0.063) (0.064)
2. Some -0.124*** -0.005 -0.098*** 0.002 -0.191*** -0.170**
(0.007) (0.044) (0.008) (0.012) (0.021) (0.067)
3. A little 0.059*** -0.007 -0.013** 0.080*** -0.079*** 0.021**
(0.021) (0.062) (0.007) (0.029) (0.006) (0.008)
4. None 0.494*** -0.006 0.596*** 0.182 0.646*** 0.297**
(0.098) (0.056) (0.055) (0.111) (0.036) (0.139)
No. of obs. 16922 16922 16922 16922 16922 16922
a
*** significant at 1% level, ** significant at 5% level, * significant at 10 % level. Clustered standard
errors in parentheses for 1773 establishments. All models include controls for education, age and
occupation.
(i.e. the positive relationship between Distrust and each of the individual centralization measures)
holds for five of these new variables. The exception is “How you do your work”, for which the
relationship with Distrust is insignificant. The instrument Inherited Distrust is highly significant in
the first stage, and the first stage F-test on the instrument is approximately 34, well above the 10
often used as a rule-of-thumb.
In order to capture the impact of Distrust on Centralization, we calculate the associated marginal
effects for each of the four possible outcomes for each of the six dependent variables (reported in
Table 8).35 As one might expect from Table 7 the marginal effects of Distrust are statistically
significant for all measures of autonomy, apart from “How you do your work”. For the other five
variables the qualitative pattern is similar: Distrust is related to a shift in probability from the
delegated outcomes (e.g. “A lot” of influence), towards centralization (“None”). The magnitudes
of the marginal effects vary markedly across the variables. Distrust has the strongest impact on
the “type of work”, “start/finish times” and “establishment organization”. The marginal effects on
“pace of work” and “decisions that affect you” are approximately half the absolute size of those for
the prior three variables. One can ex post rationalize these results in terms of conjectured technical
and coordination restrictions; it would be interesting to see these kinds of factors included in future
theory to enrich its empirical scope.
5 Concluding Comments
If a manager is willing to delegate authority to a subordinate when contracts are incomplete, she
must, in some sense, trust them to do the right thing. An employee will also, no doubt, have an
35These are the discrete changes in probability from a one unit change in Distrust, not the derivatives.
24
opinion about the trustworthiness of management. This is relevant to the design of an organization
because if an employee trusts their superiors, management can make promises with some credibility,
even in the absence of legal enforcement. Theories like relational-contracting models formalize this
idea; within a repeated-game framework, trust of management by employees enables delegation. If
feasible, such relationships/implicit contracts can enhance firm value by enhancing the effective use
of employees’ knowledge and skills.36
Our empirical analysis show a highly significant and positive relationship between delegation and
an individual employee’s trust of management. This is true for both our pooled OLS and fixed-effects
estimates. The results also suggest an endogeneity between trust and delegation. Nonetheless, the
trust-delegation relationship remains – indeed, becomes stronger – when we instrument for trust
in our IV estimates. Third, consistent with economic models of organizations, a number of other
factors are also significantly related to delegation, including employee occupation, gender and human
capital.
Finally, while our results confirm a significant relationship between trust and delegation, we do
not explain here which actions are most important in the evolution (or maintenance) of trust. A
key practical question for managers and researchers is ‘what produces the high-trust equilibrium in
firms?’. Indeed, understanding the answer to this question could help explain persistent performance
differences between seemingly similar firms (Gibbons and Henderson, 2013). Our results indicate that
trust is better understood as an individual-level relationship between an employee and management,
rather than as a group-level organizational culture. Furthermore, consistent with the burden of
past promises argument37 of Li et al. (2017), we also find trust tends to deteriorate over time. Our
analysis is just a first step, and the critical mechanism linking trust, to delegation and on to better
firm performance requires further research.
References
Acemoglu, D., P. Aghion, C. Lelarge, J. Van Reenen, and F. Zilibotti (2007). Technology, information, and the decentralization of the firm. Quarterly Journal of Economics 122 (4), 1759–1799.
Aghion, P., N. Bloom, B. Lucking, R. Sadun, and J. V. Reenen (2017). Turbulence, firm decentralization and growth in bad times. Harvard Business School Working Paper, No. 17-092.
Aghion, P., D. Fudenberg, R. Holden, T. Kunimoto, and O. Tercieux (2012). Subgame-perfect
implementation under information perturbations. The Quarterly Journal of Economics 127 (4),
1843–1881.
Aghion, P. and R. Holden (2011). Incomplete contracts and the theory of the firm: What have we
learned over the past 25 years? Journal of Economic Perspectives 25 (2), 181–197.
Aghion, P. and J. Tirole (1997). Formal and real authority in organizations. Journal of Political
Economy 105 (1), 1–29.
Algan, Y. and P. Cahuc (2010). Inherited trust and growth. The American Economic Review 100 (5),
2060–2092.
36Trust is potentially also important in facilitating communication in the absence of formal contracts. This would
be an interesting topic for future research.
37Intuitively, as circumstance change due to random shocks, the cost of keeping past promises goes up until eventually a firm defaults on a promise or, alternatively, random shocks themselves cause a breach of promise.
25
Alonso, R., W. Dessein, and N. Matouschek (2008, March). When does coordination require centralization? American Economic Review 98 (1), 145–179.
Angrist, J. D. and J.-S. Pischke (2008). Mostly harmless econometrics: An empiricist’s companion.
Princeton university press.
Arrow, K. J. (1972). Gifts and exchanges. Philosophy & Public Affairs 1 (4), 343–362.
Baker, G., R. Gibbons, and T. Hubbard (2002). Relational contracts and the theory of the firm.
Quarterly Journal of Economics 117 (1), 39–84.
Baker, G., R. Gibbons, and K. Murphy (1999). Informal authority in organizations. Journal of Law,
Economics and Organization 15 (1), 56–73.
Barrenechea-M`endez, M., P. Ort´ın-Angel, and E. Rodes (2016). Autonomy and monitoring. ´ Journal
of Economics & Management Strategy 25 (4), 911–935.
Baumeister, R. F., E. Bratslavsky, C. Finkenauer, and K. D. Vohs (2001). Bad is stronger than
good. Review of general psychology 5 (4), 323.
Berg, J., J. Dickhaut, and K. McCabe (1995). Trust, reciprocity, and social history. Games and
economic behavior 10 (1), 122–142.
Bester, H. (2009). Externalities, communication and the allocation of decision rights. Economic
Theory 41, 269 – 296.
Bidner, C. and P. Francois (2011). Cultivating trust: Norms, institutions and the implications of
scale. The Economic Journal 121 (555), 1097–1129.
Bisin, A. and T. Verdier (2001). The economics of cultural transmission and the dynamics of
preferences. Journal of Economic Theory 97 (2), 298–319.
Bisin, A. and T. Verdier (2010, November). The economics of cultural transmission and socialization.
Working Paper 16512, National Bureau of Economic Research.
Blau, F. D., L. M. Kahn, A. Y.-H. Liu, and K. L. Papps (2013). The transmission of women’s
fertility, human capital, and work orientation across immigrant generations. Journal of Population
Economics 26 (2), 405–435.
Bloom, N., R. Sadun, and J. Van Reenen (2010). Does product market competition lead firms to
decentralize? American Economic Review 100 (2), 434–438.
Bloom, N., R. Sadun, and J. Van Reenen (2012). The organization of firms across countries. Quarterly Journal of Economics 127 (4), 1663–1705.
Boedker, C., R. Vidgen, K. Meagher, J. Cogin, J. Mouritsen, and J. M. Runnalls (2011). Leadership,
Culture and Management Practices of High Performing Workplaces in Australia: High Performing
Workplace Index. Sydney: Society for Knowledge Economics.
Bolton, P. and M. Dewatripont (1994). The firm as a communication network. Quarterly Journal
of Economics 109 (4), 809–839.
Brown, S., D. Gray, J. McHardy, and K. Taylor (2015). Employee trust and workplace performance.
Journal of Economic Behavior & Organization 116, 361–378.
Butler, J. V., P. Giuliano, and L. Guiso (2016). The right amount of trust. Journal of the European
Economic Association 14 (5), 1155–1180.
Camerer, C. (2003). Behavioral game theory: Experiments in strategic interaction. Princeton University Press.
26
Colombo, M. and M. Delmastro (2004). Delegation of authority in business organizations: an
empirical test. Journal of Industrial Economics 52 (1), 53–80.
Dessein, W. (2002, October). Authority and communication in organizations. Review of Economic
Studies 69, 811–838.
Dohmen, T., A. Falk, D. Huffman, and U. Sunde (2012). The intergenerational transmission of risk
and trust attitudes. Review of Economic Studies 79 (2), 645–677.
Dreben, E. K., S. T. Fiske, and R. Hastie (1979). The independence of evaluative and item information: Impression and recall order effects in behavior-based impression formation. Journal of
Personality and Social Psychology 37 (10), 1758.
Fern´andez, R. and A. Fogli (2009, January). Culture: An empirical investigation of beliefs, work,
and fertility. American Economic Journal: Macroeconomics 1 (1), 146–77.
Fiske, S. T. (1980). Attention and weight in person perception: The impact of negative and extreme
behavior. Journal of personality and Social Psychology 38 (6), 889.
Flores-Fillol, R., S. Iranzo, and F. Mane (2017). Teamwork and delegation of decisions within the
firm. International Journal of Industrial Organization 52, 1–29.
Garicano, L. (2000). Hierarchies and the organization of knowledge in production. Journal of
Political Economy 14, 159–181.
Gibbons, R. (1998). Incentives in organizations. Journal of Economic Perspectives 12 (4), 115–132.
Gibbons, R. and R. Henderson (2013). What do managers do? In R. Gibbons and J. Roberts (Eds.),
Handbook of Organizational Economics, pp. 680–731. Princeton University Press.
Glaeser, E. L., D. I. Laibson, and C. L. Scheinkman, Jos A. Soutter (2000). Measuring trust.
Quarterly Journal of Economics 115 (3), 811–846.
Green, E. J. and R. H. Porter (1984). Noncooperative collusion under imperfect price information.
Econometrica: Journal of the Econometric Society, 87–100.
Greif, A. (1993). Contract enforceability and economic institutions in early trade: The Maghribi
traders’ coalition. The American Economic Review, 525–548.
Grossman, S. J. and O. D. Hart (1986, Aug). The costs and benefits of ownership: A theory of
vertical and lateral integration. Journal of Political Economy 4, 691–719.
Guadalupe, M., H. Li, and J. Wulf (2014). Who lives in the C-suite? Organizational structure and
the division of labor in top management. Management Science 60 (4), 824–844.
Guadalupe, M. and J. Wulf (2010, October). The flattening firm and product market competition:
The effect of trade liberalization on corporate hierarchies. American Economic Journal: Applied
Economics 2 (4), 105–27.
Guiso, L., P. Sapienza, and L. Zingales (2008). Social capital as good culture. Journal of the
European Economic Association 6 (2-3), 295–320.
Guiso, L., P. Sapienza, and L. Zingales (2015). The value of corporate culture. Journal of Financial
Economics 117 (1), 60–76. NBER Conference on the Causes and Consequences of Corporate
Culture.
Gujarati, D. (2005). Essentials of Econometrics (Third ed.). McGraw-Hill/Irwin.
Gulati, R. and J. A. Nickerson (2008). Interorganizational trust, governance choice, and exchange
performance. Organization Science 19 (5), 688708.
27
Hackman, J. R. and G. R. Oldham (1975). Development of the job diagnostic survey. Journal of
Applied Psychology 60 (2), 159–170.
Hart, O. and J. Moore (2008). Contracts as reference points. Quarterly Journal of Economics 123 (1),
1–48.
Hong, B., L. Kueng, and M.-J. Yang (2017, October). Complementarity of performance pay and
task allocation. University of Washington, Seattle.
Jensen, M. and W. Meckling (1998). Specific and general knowledge, and organizational structure.
In M. Jensen (Ed.), Foundations of Organizational Strategy. Cambridge Massachusetts: Harvard
University Press.
Kahneman, D. (2011). Thinking Fast and Slow. Farrar, Straus and Giroux.
Kleibergen, F. and R. Paap (2006). Generalized reduced rank tests using the singular value decomposition. Journal of econometrics 133 (1), 97–126.
Kleibergen, F. and M. E. Schaffer (2007, August). RANKTEST: Stata module to test the rank of a
matrix using the Kleibergen-Paap rk statistic. Statistical Software Components, Boston College
Department of Economics.
Knack, S. and P. Keefer (1997). Does social capital have an economic payoff? A cross-country
investigation. The Quarterly journal of economics, 1251–1288.
Kreps, D. (1990). Corporate culture and economic theory. In J. Alt and K. Shepsle (Eds.), Perspectives on Positive Political Economy, pp. 90–143. Cambridge University Press.
La Porta, R. et al. (1997). Trust in large organizations. American Economic Review 87 (2), 333–38.
Levin, J. (2003). Relational incentive contracts. American Economic Review 93 (3), 835–857.
Li, J., N. Matouschek, and M. Powell (2017). Power dynamics in organizations. American Economic
Journal: Microeconomics 9 (1), 217–41.
Macchiavello, R. and A. Morjaria (2015). The value of relationships: evidence from a supply shock
to kenyan rose exports. American Economic Review 105 (9), 2911–2945.
Malcomson, J. M. (2013). Relational incentive contracts. In R. Gibbons and J. Roberts (Eds.),
Handbook of Organizational Economics, Chapter 25, pp. 1014–1065. Princeton University Press.
Mayer, R. C., J. H. Davis, and F. D. Schoorman (1995). An integrative model of organizational
trust. Academy of Management Review 20 (3), 709–734.
McElheran, K. (2014). Delegation in multi-establishment firms: Evidence from I.T. purchasing.
Journal of Economics & Management Strategy 23 (2), 225–258.
McEvily, B. and M. Tortoriello (2011). Measuring trust in organisational research: Review and
recommendations. Journal of Trust Research 1 (1), 23–63.
Meagher, K., H. Orbay, and T. V. Zandt (2004). Market contingent managerial hierarchies. University of New South Wales.
Meagher, K. J. (2003). Generalizing incentives and loss of control in an optimal hierarchy: the role
of information technology. Economics Letters 78 (2), 273–280.
Meagher, K. J. and A. Wait (2014). Delegation of decisions about change in organizations: The roles
of competition, trade, uncertainty and scale. Journal of Law, Economics and Organization 30 (4),
709–733.
28
Morehead, A., M. Steele, M. Alexander, K. Stephen, and L. Duffin (1997). Changes at work: the 1995
Australian Workplace Industrial Relations Survey. South Melbourne: Addison Welsey Longman
Australia.
Nunn, N. and L. Wantchekon (2011). The slave trade and the origins of mistrust in Africa. American
Economics Review 101, 3221–3252.
Pratto, F. and O. P. John (1991). Automatic vigilance: the attention-grabbing power of negative
social information. Journal of personality and social psychology 61 (3), 380.
Prendergast, C. (2002). The tenuous trade-off between risk and incentives. Journal of Political
Economy 110 (5), 1071–1102.
Radner, R. (1993). The organization of decentralized information processing. Econometrica 61 (5),
1109–1146.
Rajan, R. G. and J. Wulf (2006, November). The flattening firm: Evidence from panel data on the
changing nature of corporate hierarchies. The Review of Economics and Statistics 88 (4), 759–773.
Rossi-Hansberg, E., Caliendo, and Monte (2014). The anatomy of French production hierarchies.
Journal of Political Economy 123 (4), 809–852.
Rossi-Hansberg, E. and L. Garicano (2012, March). Organizing growth. Journal of Economic
Theory 147 (2), 623–656.
Rousseau, D., S. Sitkin, R. Burt, and C. Camerer (1998). Not so different after all: A cross-discipline
view of trust. Academy of Management Review 23 (3), 393–404.
Rozin, P. and E. B. Royzman (2001). Negativity bias, negativity dominance, and contagion. Personality and social psychology review 5 (4), 296–320.
Sako, M. and S. Helper (1998). Determinants of trust in supplier relations: Evidence from the
automotive industry in japan and the united states. Journal of Economic Organization and
Behavior 42, 387–417.
Shockley-Zalabak, P., S. Morreale, and M. Hackman (2010). Building the high-trust organization:
Strategies for supporting five key dimensions of trust. San Fransisco CA: John Wiley & Sons.
Skowronski, J. J. and D. E. Carlston (1987). Social judgment and social memory: The role of
cue diagnosticity in negativity, positivity, and extremity biases. Journal of personality and social
psychology 52 (4), 689.
Spreitzer, G. (1995). Psychological empowerment in the workplace: Dimensions, measurement, and
validation. Academy of Management Journal 38 (5), 1442–1465.
Stinchcombe, A. L. (1990). Information and Organizations. Number 19 in California Series on Social
Change and Political Economy. University of California Press.
Stock, J. and M. Yogo (2005). Testing for weak instruments in linear IV regression. In D. Andrews
and J. Stock (Eds.), Identification and Inference for Econometric Models: Essays in Honor of
Thomas Rothenberg, pp. 80–108. Cambridge University Press.
Stock, J. H. and M. W. Watson (2015). Introduction to econometrics. Pearson.
Tabellini, G. (2010). Culture and institutions: economic development in the regions of europe.
Journal of the European Economic Association 8 (4), 677–716.
Thomas, K. W. (1990). Cognitive elements of empowerment: An “interpretive” model of intrinsic
task motivation. Academy of Management Review 15 (4), 666–681.
29
Thorelli, H. (1986). Networks: Between markets and hierarchies. Strategic Management Journal 18,
509–533.
Uslaner, E. M. (2008). Where you stand depends upon where your Grandparents sat: The inheritability of generalized trust. Public Opinion Quarterly 72 (4), 725–740.
Van Zandt, T. (1999). Real-time decentralized information processing as a model of organizations
with boundedly rational agents. Review of Economic Studies 66, 633–658.
Williams, D. (2013). The 7 Non-Negotiables of Winning: Tying Soft Traits to Hard Results. Hoboken
NJ: John Wiley & Sons.
Williamson, O. E. (1985). The Economic Institutions of Capitalism: Firms, Markets, Relational
Contracting. New York: The Free Press.
Wooldridge, J. M. (2000). Introductory Econometrics: A Modern Approach. Thomson Learning.
Zabojnik, J. (2002). Centralized and decentralized decision making in organizations. Journal of
Labor Economics 20 (1), 1–22.
A Other empirical results
Table 9: Correlations between Education, Distrust and Centralization (N = 16922)
Variable Centralization Distrust
Year 10 0.062 -0.031
Year 12 0.114 -0.0127
Vocational 0.030 0.0010
Skilled trade 0.018 0.0763
Associate Diploma 0.045 0.0121
Undergraduate degree -0.038 -0.0199
Postgraduate degree -0.109 -0.0046
Other -0.137 -0.0049
Notes: a. Source AWIRS 95.
Table 10: Correlations between Occupation, Disrust and Centralization (N = 16922)
Variable Centralization Distrust
Laborer 0.164 -0.1218
Machine operator 0.114 -0.1023
Sales & Personal service 0.088 -0.0965
Clerks 0.003 -0.1333
Tradesperson 0.062 -0.1010
Para-professional 0.015 -0.0577
Professional -0.134 0.4384
Manager -0.315 0.1355
Other -0.007 -0.0175
Notes: a. Source AWIRS 95.
30

---

Get Professional Assignment Help Cheaply

Are you busy and do not have time to handle your assignment? Are you scared that your paper will not make the grade? Do you have responsibilities that may hinder you from turning in your assignment on time? Are you tired and can barely handle your assignment? Are your grades inconsistent?

Whichever your reason is, it is valid! You can get professional academic help from our service at affordable rates. We have a team of professional academic writers who can handle all your assignments.

Why Choose Our Academic Writing Service?

• Plagiarism free papers
• Timely delivery
• Any deadline
• Skilled, Experienced Native English Writers
• Subject-relevant academic writer
• Adherence to paper instructions
• Ability to tackle bulk assignments
• Reasonable prices
• 24/7 Customer Support
• Get superb grades consistently

Online Academic Help With Different Subjects

Literature

Students barely have time to read. We got you! Have your literature essay or book review written without having the hassle of reading the book. You can get your literature paper custom-written for you by our literature specialists.

Finance

Do you struggle with finance? No need to torture yourself if finance is not your cup of tea. You can order your finance paper from our academic writing service and get 100% original work from competent finance experts.

Computer science

Computer science is a tough subject. Fortunately, our computer science experts are up to the match. No need to stress and have sleepless nights. Our academic writers will tackle all your computer science assignments and deliver them on time. Let us handle all your python, java, ruby, JavaScript, php , C+ assignments!

Psychology

While psychology may be an interesting subject, you may lack sufficient time to handle your assignments. Don’t despair; by using our academic writing service, you can be assured of perfect grades. Moreover, your grades will be consistent.

Engineering

Engineering is quite a demanding subject. Students face a lot of pressure and barely have enough time to do what they love to do. Our academic writing service got you covered! Our engineering specialists follow the paper instructions and ensure timely delivery of the paper.

Nursing

In the nursing course, you may have difficulties with literature reviews, annotated bibliographies, critical essays, and other assignments. Our nursing assignment writers will offer you professional nursing paper help at low prices.

Sociology

Truth be told, sociology papers can be quite exhausting. Our academic writing service relieves you of fatigue, pressure, and stress. You can relax and have peace of mind as our academic writers handle your sociology assignment.

Business

We take pride in having some of the best business writers in the industry. Our business writers have a lot of experience in the field. They are reliable, and you can be assured of a high-grade paper. They are able to handle business papers of any subject, length, deadline, and difficulty!

Statistics

We boast of having some of the most experienced statistics experts in the industry. Our statistics experts have diverse skills, expertise, and knowledge to handle any kind of assignment. They have access to all kinds of software to get your assignment done.

Law

Writing a law essay may prove to be an insurmountable obstacle, especially when you need to know the peculiarities of the legislative framework. Take advantage of our top-notch law specialists and get superb grades and 100% satisfaction.

What discipline/subjects do you deal in?

We have highlighted some of the most popular subjects we handle above. Those are just a tip of the iceberg. We deal in all academic disciplines since our writers are as diverse. They have been drawn from across all disciplines, and orders are assigned to those writers believed to be the best in the field. In a nutshell, there is no task we cannot handle; all you need to do is place your order with us. As long as your instructions are clear, just trust we shall deliver irrespective of the discipline.

Are your writers competent enough to handle my paper?

Our essay writers are graduates with bachelor's, masters, Ph.D., and doctorate degrees in various subjects. The minimum requirement to be an essay writer with our essay writing service is to have a college degree. All our academic writers have a minimum of two years of academic writing. We have a stringent recruitment process to ensure that we get only the most competent essay writers in the industry. We also ensure that the writers are handsomely compensated for their value. The majority of our writers are native English speakers. As such, the fluency of language and grammar is impeccable.

What if I don’t like the paper?

There is a very low likelihood that you won’t like the paper.

Reasons being:

• When assigning your order, we match the paper’s discipline with the writer’s field/specialization. Since all our writers are graduates, we match the paper’s subject with the field the writer studied. For instance, if it’s a nursing paper, only a nursing graduate and writer will handle it. Furthermore, all our writers have academic writing experience and top-notch research skills.
• We have a quality assurance that reviews the paper before it gets to you. As such, we ensure that you get a paper that meets the required standard and will most definitely make the grade.

In the event that you don’t like your paper:

• The writer will revise the paper up to your pleasing. You have unlimited revisions. You simply need to highlight what specifically you don’t like about the paper, and the writer will make the amendments. The paper will be revised until you are satisfied. Revisions are free of charge
• We will have a different writer write the paper from scratch.
• Last resort, if the above does not work, we will refund your money.

Will the professor find out I didn’t write the paper myself?

Not at all. All papers are written from scratch. There is no way your tutor or instructor will realize that you did not write the paper yourself. In fact, we recommend using our assignment help services for consistent results.

What if the paper is plagiarized?

We check all papers for plagiarism before we submit them. We use powerful plagiarism checking software such as SafeAssign, LopesWrite, and Turnitin. We also upload the plagiarism report so that you can review it. We understand that plagiarism is academic suicide. We would not take the risk of submitting plagiarized work and jeopardize your academic journey. Furthermore, we do not sell or use prewritten papers, and each paper is written from scratch.

When will I get my paper?

You determine when you get the paper by setting the deadline when placing the order. All papers are delivered within the deadline. We are well aware that we operate in a time-sensitive industry. As such, we have laid out strategies to ensure that the client receives the paper on time and they never miss the deadline. We understand that papers that are submitted late have some points deducted. We do not want you to miss any points due to late submission. We work on beating deadlines by huge margins in order to ensure that you have ample time to review the paper before you submit it.

Will anyone find out that I used your services?

We have a privacy and confidentiality policy that guides our work. We NEVER share any customer information with third parties. Noone will ever know that you used our assignment help services. It’s only between you and us. We are bound by our policies to protect the customer’s identity and information. All your information, such as your names, phone number, email, order information, and so on, are protected. We have robust security systems that ensure that your data is protected. Hacking our systems is close to impossible, and it has never happened.

How our Assignment Help Service Works

1. Place an order

You fill all the paper instructions in the order form. Make sure you include all the helpful materials so that our academic writers can deliver the perfect paper. It will also help to eliminate unnecessary revisions.

2. Pay for the order

Proceed to pay for the paper so that it can be assigned to one of our expert academic writers. The paper subject is matched with the writer’s area of specialization.

3. Track the progress

You communicate with the writer and know about the progress of the paper. The client can ask the writer for drafts of the paper. The client can upload extra material and include additional instructions from the lecturer. Receive a paper.

4. Download the paper

The paper is sent to your email and uploaded to your personal account. You also get a plagiarism report attached to your paper.

GET A PERFECT SCORE!!!

---