Administrative data and long-term trends in child maltreatment: the prospects and pitfalls

Research on child maltreatment in high-income countries has established several key facts. First, child maltreatment and contact with child protective services are common.1Wildeman C The incredibly credible prevalence of child protective services contact in New Zealand and the United States.Am J Public Health. 2018; 108: 438-439Crossref PubMed Scopus (5) Google Scholar, 2Gilbert R Widom CS Browne K Fergusson D Webb E Janson S Burden and consequences of child maltreatment in high-income countries.Lancet. 2009; 373: 68-81Summary Full Text Full Text PDF PubMed Scopus (2461) Google Scholar Second, child maltreatment does children severe harm and has serious social costs.2Gilbert R Widom CS Browne K Fergusson D Webb E Janson S Burden and consequences of child maltreatment in high-income countries.Lancet. 2009; 373: 68-81Summary Full Text Full Text PDF PubMed Scopus (2461) Google Scholar, 3Fang X Brown DS Florence CS Mercy JA The economic burden of child maltreatment in the United States and implications for prevention.Child Abuse Negl. 2012; 36: 156-165Crossref PubMed Scopus (658) Google Scholar However, data limitations have made it difficult to document whether child maltreatment is increasing or decreasing, especially when viewed in a larger historical context.4Gilbert R Fluke J O'Donnell M et al.Child maltreatment: variation in trends and policies in six developed countries.Lancet. 2012; 379: 758-772Summary Full Text Full Text PDF PubMed Scopus (216) Google Scholar In The Lancet Public Health, Michelle Degli Esposti and colleagues5Degli Esposti M Humphreys DK Jenkins BM et al.Long-term trends in child maltreatment in England and Wales from official record data, 1858–2016: an observational, time-series analysis.Lancet Public Health. 2019; 4: e148-e158Summary Full Text Full Text PDF PubMed Scopus (17) Google Scholar make a laudable contribution to this area of research with an observational time-series analysis, using official government and charity record data to study trends in child maltreatment in England and Wales over a 158-year period. That contribution rests on two strategies. The first is to compile and compare data on multiple measures of child abuse and neglect. No single measure of abuse and neglect can be decisive because each captures only an aspect of this heterogeneous construct and depends on a combination of underlying behaviour, administrative effort, and historically shifting norms. By triangulating across measures, one can better assess the validity of each. The second strategy is to construct a lengthy time-series analysis, which the authors make publicly available in a new database, iCoverT. As a result, it is possible to know whether alternative measures of maltreatment covary, and whether trends over the past decade are continuations of historical patterns, meaningful anomalies, or random noise. These strategies enable the authors to arrive at several novel findings.5Degli Esposti M Humphreys DK Jenkins BM et al.Long-term trends in child maltreatment in England and Wales from official record data, 1858–2016: an observational, time-series analysis.Lancet Public Health. 2019; 4: e148-e158Summary Full Text Full Text PDF PubMed Scopus (17) Google Scholar First, fatal violence against children in England and Wales has steadily decreased for the past 150 years (90% decrease in child death by homicide or assault between 1858 and 2016); this trend is most prominent among children younger than 1 year. Second, in the 21st century, registrations with child protection services have proliferated (182% increase between 1988 and 2016), driven increasingly by neglect and emotional abuse rather than physical abuse. Corresponding increases in criminal convictions help rule out the possibility that increased registrations are solely a function of increased effort by child protection services. The Article nevertheless highlights some of the characteristic challenges of using administrative data. In a particularly impressive effort, the authors show that formal changes in definitions and recording practices do not meaningfully bias their findings—an important test that historical analyses such as these often do not take into account. However, the more important changes might be informal ones—a possibility the authors acknowledge only in passing. Over the past several centuries, children have come to be seen as more distinctive, precious, and sensitive.6Aries P Centuries of childhood: a social history of family life. Random House, New York1962: 411-418Google Scholar, 7Zelizer VA Pricing the priceless child: the changing social value of children. Basic Books, New York1985: 3-21Google Scholar These changing views have shaped how their treatment is perceived. Similar problems characterise other administrative data on families, such as divorces for spousal cruelty.8Griswold RL The evolution of the doctrine of mental cruelty in Victorian American divorce, 1790–1900.J Soc Hist. 1986; 20: 127-148Crossref PubMed Scopus (18) Google Scholar Therefore, rather than just measuring a stable set of objective childhood experiences, long-term administrative data on maltreatment capture a basket of socially undesirable behaviours whose contents change over time. These nuances recommend caution in interpretation. But they also recommend extension of the authors' strategy of comparing multiple data sources. For example, the analysis rightly treats homicide data as an objective complement to more ambiguous data on child registrations for physical abuse. But, because neglect now makes up a substantial plurality of registrations, coupling maltreatment data with data on malnutrition or accidental deaths could help researchers understand the meaning of increasing neglect registrations amid changing conceptions of neglect.9Swift KJ Manufacturing ‘bad mothers’: a critical perspective on child neglect. University of Toronto Press, Toronto1995: 37-63Crossref Google Scholar Likewise, data on per capita expenditure on child protection services could give important context for overall increases in registrations. The authors make one error of historical reasoning in this otherwise excellent analysis. When summarising long-term trends, they cite proportional changes in incidences over the period of analysis. Because their data are left-censored by availability, and because many observed trends are non-monotonic (a finding in its own right), these changes are sometimes arbitrary—not only the magnitude but also the direction of observed trends depends on when the data come into existence. Furthermore, the historical availability of alterative measures varies widely, so comparison of total changes over differing periods of analysis can be misleading. As the rapid digitisation of historical data opens exciting new frontiers in research, it will be important to remember the consequences of missing data and to compare like for like when analysing trends. Through the innovative use of multiple sources of administrative data, Esposti and colleagues5Degli Esposti M Humphreys DK Jenkins BM et al.Long-term trends in child maltreatment in England and Wales from official record data, 1858–2016: an observational, time-series analysis.Lancet Public Health. 2019; 4: e148-e158Summary Full Text Full Text PDF PubMed Scopus (17) Google Scholar generate some definitive conclusions about long-term decreases in severe violence against children. If their findings about trends since the turn of the century are less conclusive, their approach nevertheless charts a careful path for navigating the thorny problem of measuring trends in child abuse and neglect, a path that others should emulate. We declare no competing interests. Long-term trends in child maltreatment in England and Wales, 1858–2016: an observational, time-series analysisAlthough long-term trends have decreased, child maltreatment remains a major public health problem in England and Wales. Further research is needed to establish whether adolescents are a particularly vulnerable age group and whether neglect and emotional abuse are increasing. Future child protection policies and practices should respond to these areas of growing need. Full-Text PDF Open Access

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.