Triadic signatures of global human mobility networks

Abstract

Global refugee and migrant flows form complex networks with serious consequences for both sending and receiving countries as well as those in between. While several basic network properties of these networks have been documented, their finer structural character remains under-studied. One such structure is the triad significance profile (TSP). In this study, the TSPs of global refugee and migrant flow networks are assessed. Results indicate that the migrant flow network's size and TSP remain stable over the years; its TSP shares patterns with social networks such as trade networks. In contrast, the refugee network has been more dynamic and structurally unstable; its TSP shares patterns with networks in the information-processing superfamily, which includes many biological networks. Our findings demonstrate commonality between migrant and social networks as well as between refugee and biological networks, pointing to possible interdisciplinary collaboration-e.g., application of biological network theories to refugee network dynamics-, potentially furthering theoretical development with respect to both network theory and theories on human mobility.

Fig 1. More flows, more edges? Not quite.

The relationship between the sum of all refugee flows and the number of edges in the refugee flow network: (A) the original flow data, and (B) after removing edges with fewer than 100 refugees per year. The relationship is complex and exhibits significant variation, with certain years showing increased flows despite a reduced number of edges.

Fig 2. Shuffling for significance.

Schematic diagram of the rewiring process to create a randomized network.

Fig 3. Trends of the low-order properties of the human mobility networks.

(A) Numbers of countries involved with the human mobility networks composed of edges with 100 migrants/refugees or more per year; (B), (C), and (D) Edge compositions of the migrant, annual refugee, and 5-year aggregate refugee flow networks, respectively: diamonds = density of mutual edges (m), triangle = density of asymmetric edges a, and thick line = overall density (m + a). Here density is calculated as the fraction of the total number of possible N(N − 1)/2 node pairs, with N being the number of nodes, that are occupied by certain types of edges (mutual, asymmetric, or null).

Fig 4. Fragmented nature of the refugee flow networks.

(A) The numbers of disconnected components of the refugee flow networks across the years; and (B) The size (number of nodes) of the biggest component in each year.

Fig 5. Triadic signatures of the human mobility networks.

(A) UNDESA migrants; (B) 5-year aggregate UNHCR refugees (only largest components); and (C) annual UNHCR refugees (only largest components). For each box plot, central red mark = median; bottom edge = 25^th percentile; top edge = 75^th percentile; whiskers extend to the most extreme data points not considered outliers; red plus signs represent data points considered outliers [16].

Fig 6. Example triads that shaped the TSP of the UNHCR refugee flow network.

For each year, we considered triads of the strongly over-represented types; for each of these over-represented types, the triad with the greatest volume was shown. The percentage of total refugee flows contained in the largest network component (from which the TSP is derived) is shown in the parentheses: for the 2001–2005 and 2006–2010 periods where the percentages were relatively low compared to the other periods, the second biggest components contain 21% and 23% of total refugee flows, respectively—that is, the largest components still contain significantly more refugee flows. The country acronyms are defined as follows: BDI = Burundi; CAF = Central African Republic; COD = Democratic Republic of the Congo; CMR = Cameroon; ETH = Ethiopia; IRQ = Iraq; KEN = Kenya; LBN = Lebanon; RWA = Rwanda; SDN = Sudan; SOM = Somalia; SSD = South Sudan; SYR = Syria; TCD = Chad; TZA = Tanzania; UGA = Uganda.