Modelling supply chain adaptation for disruptions: An empirically grounded complex adaptive systems approach

• DOI: http://doi.org/10.1002/joom.1009
• Author: Jennifer V. Blackhurst,Zhiya Zuo,Kang Zhao
• Date: 01 March 2019
• Published date: 01 March 2019

ORIGINAL ARTICLE

Modelling supply chain adaptation for disruptions:

An empirically grounded complex adaptive systems approach

Kang Zhao

1

| Zhiya Zuo

2

| Jennifer V. Blackhurst

1

1

Department of Management Sciences,

Tippie College of Business, University of

Iowa, Iowa City, Iowa

2

Department of Information Systems,

City University of Hong Kong,

Kowloon Tong, Hong Kong

Correspondence

Jennifer V. Blackhurst, Department of

Management Sciences, Tippie College of

Business, University of Iowa

Iowa City, IA 52240.

Email: jennifer-blackhurst@uiowa.edu

Handling Editors: Anand Nair and Felix

Reed-Tsochas

Abstract

Through the development and usage of an agent-based model, this article investigates

firms' adaptive strategies against disruptions in a supply chain network. Viewing sup-

ply chain networks as complex adaptive systems, we first construct and analyze a real-

worldsupplychainnetworkamong2,971firms spanning 90 industry sectors. We then

develop an agent-based simulation to show how the model of firms' adaptive behaviors

can leverage competition relationships within a supply chain network. The simulation

also models how disruptions propagate in the supply chain network through cascading

failures. With the simulation, we seek to understand if a firm's adaptive behaviors can

reduce the impact of disruptions in supply chain networks. Therefore, we propose,

evaluate, and analyze two types of adaptive strategies a firm can leverage to reduce the

negative effects of supply chain network disruptions. First, we deploy in our model a

reactive strategy, which restructures the network in response to a disruption event

among first-tier suppliers. Next, we develop and propose proactive strategies, which

are used when a distant disruption is observed but has not yet hit the focal firm. We

discuss the implications related to how and when firms can improve their resilience

against supply disruptions by leveraging adaptive strategies.

KEYWORDS

agent-based models, complex networks, resilience, supply chain disruptions

1|INTRODUCTION

Due to the complexity, uncertainty, and interdependence of

today's supply chains, there is an increased risk of loss in the

supply chain network due to a disruption event (Bode &

Wagner, 2015; Bode, Wagner, Petersen, & Ellram, 2011;

Kamalahmadi & Parast, 2016). A disruption in a supply chain

network is defined as an event that disrupts the flow of

goods or services (Craighead, Blackhurst, Rungtusanatham, &

Handfield, 2007). Losses stemming from supply chain network

disruptions may manifest as financial loss, a loss in operational

performance, and even a loss of market position (Hendricks &

Singhal, 2003; Hendricks & Singhal, 2005; Wagner & Bode,

2008). Moreover, because of the interconnected nature of

supply chain networks, a disruption may propagate and cascade

through the supply chain (Fiksel, Polyviou, Croxton, & Pettit,

2015; Hearnshaw & Wilson, 2013), with increasing magnitude

or severity of impact (Van der Vegt, Essens, Wahlstrom, &

George, 2015). In other words, a disruption may not originate

from the focal firm's immediate suppliers but rather elsewhere

in the network (Blackhurst, Craighead, Elkins, & Handfield,

2005; Kim, Chen, & Linderman, 2015). A lack of understand-

ing of how the supply chain network is structured may exacer-

bate the impact of disruptions and inadvertently allow

disruptions to propagate (Kim et al., 2015). Managers of real-

world supply chains find the cascading effect or propagation of

a disruption difficult to understand (Fiksel et al., 2015). The

ability to restructure the supply chain in the face of changing

conditions is critical to maintain continuity of supply chain

DOI: 10.1002/joom.1009

190 © 2019 Association for Supply Chain Management, Inc. wileyonlinelibrary.com/journal/joom J Oper Manag. 2019;65:190–212.

performance (Hearnshaw & Wilson, 2013). Flows of materials

within the supply chain network need to be redirected and

structures need to be adapted to allow for continuity in opera-

tions. As such, there have been calls to examine the structure

of supply chain networks and determine the ability of the net-

work to adapt in the face of supply chain disruptions

(Hearnshaw & Wilson, 2013; Kim et al., 2015; Van der Vegt

et al., 2015).

In this study, we view a supply chain network as a complex

adaptive system (CAS) (Choi, Dooley, & Rungtusanatham,

2001) where, in the face of a disruption, firms connected in a

complex network have the ability to adapt and restructure their

connections. The CAS framework provides a useful theoretical

foundation for this study (Anderson, 1999; Choi et al., 2001;

Thompson, 1967) as firms in a supply chain operate as an

interconnected network in a dynamic environment (Blackhurst,

Dunn, & Craighead, 2011; Bode et al., 2011; Kim, Choi,

Yan, & Dooley, 2011). Therefore, even a small change at one

node in the chain can cause a disruption to spread, impacting

other nodes in the chain (Craighead et al., 2007). We posit that

firms in a supply chain constitute self-organizing networks. In

addition, some supply chains can be adaptive or resilient.

When hit with a disruption, they can adapt or restructure them-

selves to reach a desirable state (back to the original state, an

equivalent state, or better) (Ambulkar, Blackhurst, & Grawe,

2015). In viewing supply chain networks as an adaptive sys-

tems, the ability to adapt and restructure is critical for minimiz-

ing losses from disruptions (Ambulkar et al., 2015; Hearnshaw

& Wilson, 2013). The effectiveness of adaptive restructuring

strategies in improving network resilience after node removal

has been illustrated in other complex systems, such as food

webs (Staniczenko, Lewis, Jones, & Reed-Tsochas, 2010). In

addition, Nair and Vidal (2011) noted that network topology is

an important factor with regards to spreading disruptions.

However, recent research on resilience to supply chain disrup-

tions has not fully incorporated the role of network structures

(Kim et al., 2011) and lacks a clear understanding of disrup-

tions and their impact at a network level (Kim et al., 2015). In

other words, understanding how disruptions impact multiple

tiers in a supply chain and how the structure of the network

may play a role in this impact is lacking. In order to address

these gaps in the research, we seek to answer the following

research question:

How can firms leverage different types of adap-

tive strategies in the supply network to improve

resilience against supply disruptions?

Inspired by both supply chain management and network

science literatures on rewiring edges (Watts & Strogatz, 1998;

Zhao, Kumar, & Yen, 2011), our study presents and examines

two types of adaptive strategies to restructure a supply chain

network: (a) a reactive strategy, which restructures the network

in response to a disruption event among first-tier suppliers. In

other words, reactive strategies are used when an immediate

supplier of a focal firm fails. Next, we develop and propose

(b) proactive strategies. These strategies focus on restructuring

the network after observing a distant firm failure (beyond first

tier) in order to avoid possible disruptions to the focal firm.

Representing a forward-looking approach, proactive strategies

are in anticipation of a disruption (which has already occurred

in another part of the network) hitting the focal firm and will

identify the weakest spot specific to the disrupted distant firm

in the network.

In order to study firms' adaptive strategies that improve their

resilience to supply chain disruptions, this study develops

agent-based simulations based on large-scale real-world supply

networks. Our modeling of adaptive behaviors incorporates the

structure of both supply chain networks (which connect partner

firms in the supply chain) and competition networks (which

connect competing firms in the supply chain) so that we can

investigate how competition relationships among firms in a

supply chain network can be exploited to develop resilience

against disruptions (in Sections 4 and 5.1). The two networks

are again used to model and analyze firms' proactive strategies

(in Section 5.2) including factors related to the effectiveness of

proactive strategies (in Section 5.3).

This research proceeds in four steps: First, we collect data

of 2,971 firms from 90 industries to construct a large-scale

supply chain network among these firms, along with an

accompanying competition network. The data was collected

through scraping a database for information on firms includ-

ing their financial data as well as relationship data among

firms. We reveal the complex structural properties of these

networks and show a firm's partnership and competition with

others are interweaved. Second, we design agent-based sim-

ulation models for firms' reactive strategies in this complex

system and the propagation of disruption impact. Third, we

use the models to evaluate the impact of disruptions and

illustrate the effectiveness of reactive behaviors in reducing

the impact of disruptions. Fourth, we propose, evaluate, and

analyze proactive strategies that firms can use to improve

their supply chain resilience against distant disruptions.

This study makes a number of important contributions to

the understanding of supply chain networks. First, our

agent-based model (ABM) leverages structures of both real-

world supply chain and competition networks as well as firm

attributes to realistically model key components of complex-

ity in supply chain networks, namely the propagation of a

disruption in the supply chain and firms' adaptive behaviors

to manage disruption risk. The use of competition networks

opens interesting possibilities to not only handle disruptions

more effectively but also to gain advantage in the market by

leveraging visibility of relationships and structures within

ZHAO ET AL.191