Imagine a young person who has stopped attending school regularly. From the perspective of the education system, the problem may first appear as absenteeism. A school records missed days, contacts the family and perhaps refers the student to a wellbeing officer or attendance program. But suppose the reason is not primarily school. The family has lost secure housing and is moving between relatives. A parent has lost work. Conflict at home has increased. The young person is sleeping badly, travelling further to school and spending less time with friends. Anxiety rises. Eventually there may be contact with mental health services, Centrelink, homelessness services or police. Government may now see several problems, but the young person experiences one life.
This is not an abstract example. In 2024–25, specialist homelessness services assisted around 40,500 Australians aged 12 to 24 who were receiving support alone. Seventy-two per cent were not enrolled in formal education or training. Their needs commonly extended beyond accommodation into mental health, employment, education, transport and income support. The administrative categories are real, but so are the relationships between them. A young person does not become an education case on Monday, a housing case on Tuesday and a mental health case on Wednesday. The same pressures travel with them through each institution.
Modern government is necessarily organised by responsibility. Health departments manage health. Education departments manage education. Housing agencies manage housing. Justice agencies manage offending and public safety. Treasuries manage budgets and economic settings. Complex societies require specialist knowledge, professional capability and clear lines of accountability. The difficulty is that the lives governments are trying to improve do not divide themselves according to administrative boundaries. Housing insecurity can affect school attendance; school disengagement can affect employment; employment affects income and housing stability; family stress affects health; poor health affects participation; contact with the justice system can further narrow employment and housing options. Once several of these conditions begin interacting, it becomes increasingly difficult to identify one original cause or one department capable of fixing the outcome.
The Institute for Economics and Peace approaches this problem through systems thinking. Its 2026 Positive Peace Report argues that societies are best understood as networks of relationships, institutions, resources and behaviours whose combined effects cannot be adequately explained by examining each component separately. In its Positive Peace model, eight broad areas associated with peaceful societies operate as an interconnected system. No individual Pillar is as strongly associated with peacefulness as the combined whole, and progress in one area can be strengthened or constrained by conditions elsewhere.
This challenges one of the most persistent habits in public policy: the search for an isolated cause attached to an isolated problem. Crime rises, so attention turns to police numbers and penalties. Hospital demand increases, so governments look for beds and staff. Housing becomes unaffordable, so the debate concentrates on supply. School attendance falls, so programs target attendance. Each response may be sensible, and sometimes the direct response is exactly what is needed. The difficulty begins when the visible problem is only one expression of a larger system. The Positive Peace Report contrasts this conventional cause-and-effect approach with systems thinking, which directs attention towards relationships, flows and interactions. It argues that policies aimed at an assumed cause can produce disappointing results when wider dynamics are poorly understood, including consequences in parts of the system the intervention was never designed to affect.
Systems thinking does not mean that every problem is infinitely complicated or that government should delay action until every relationship has been mapped. The useful question is narrower: which relationships are actually shaping the outcome we are trying to change? If unstable housing is the main reason a student has disengaged from school, an attendance program may address the symptom while leaving the dominant pressure untouched. If violence at home is central, housing assistance alone may not be enough. If repeated moves have severed friendships and weakened connection to school, stabilising accommodation may help without automatically restoring participation. The intervention can be worthwhile without being sufficient.
The same pattern appears far beyond disadvantage. Transport policy affects access to employment and services. Planning affects housing, infrastructure and local economies. Environmental shocks affect health, insurance, migration and regional viability. Information policy affects public health, democratic participation and institutional trust. Economic policy affects family stability and confidence in the future. Police practices affect both public safety and legitimacy. Each area requires its own expertise, but the consequences regularly cross portfolio boundaries.
Public institutions are generally better at measuring what happens inside those boundaries than what happens between them. An agency can report how many houses were built, how many patients were treated or how many people completed a program. Those measures are necessary, but they become less informative when the more important question is whether an intervention changed the conditions generating the problem. A housing program might meet its construction target while the households under greatest pressure remain unable to access the new stock. A youth employment program might place participants into jobs while transport, unstable housing or poor health make those placements difficult to sustain. A policing strategy might reduce offending in one location while moving it elsewhere. None of these possibilities proves that the original policy failed. They reveal why the unit of analysis matters. If we measure only the program, we may miss the system.
Peace becomes particularly interesting when viewed this way. It is often understood too narrowly as the absence of war or violence, but that definition is incomplete for a society such as Australia. A more useful definition is the capacity of a society to manage disagreement, competing interests, pressure and change without escalating into destructive conflict or violence. That definition deliberately leaves room for conflict. A peaceful society is not one without protest, political opposition, industrial action, litigation, criticism or public anger. Those forms of contest are often necessary in a democracy because they expose injustice, challenge power and allow different interests to negotiate change. The relevant distinction is not between conflict and peace, but between conflict that remains legitimate, bounded and non-violent and conflict that becomes coercive, dehumanising or destructive.
IEP’s concept of Positive Peace is useful here because it looks beyond the current level of violence to the attitudes, institutions and structures associated with societies that are better able to sustain peaceful outcomes over time. The Positive Peace Index was developed from statistical analysis of more than 20,000 data series and covers 163 countries, representing 99.7 per cent of the world’s population. Its factors include government effectiveness, corruption, human capital, resource distribution, information, economic conditions and acceptance of the rights of others. These should not be treated as ingredients in a formula. Many relationships operate in both directions. Strong institutions can support peaceful societies, while peaceful conditions also make institutions easier to govern. Trust can strengthen cooperation, while competent institutions can generate trust. Economic resilience may reduce social pressure, while stability itself encourages investment.
Nor should every desirable social outcome be relabelled as peace policy. Health, education, housing and economic opportunity have value in their own right. They become relevant to peace when they influence resilience, exclusion, legitimacy, grievance or the ability of people and institutions to manage pressure without destructive escalation. Without that boundary, peace can expand until it simply becomes another word for a well-functioning society.
Australia begins from a position of considerable strength. It ranks eighth in the 2026 Positive Peace Index. The relevant question is not how to rescue a failed system, but how to understand and preserve the conditions that allow a successful one to remain resilient. Governments already recognise the need for cross-portfolio thinking in areas such as emergency management and national security because disasters and threats do not respect departmental boundaries. Yet much social policy still begins from the organisational chart. A problem is assigned to the agency with the closest mandate, funding follows, performance measures are created and the intervention is evaluated. If the underlying problem lives partly in the relationships between housing, education, health, employment, justice and community life, however, no individual department can fully own the outcome.
This creates an accountability paradox. Clear responsibility is essential for government to function, but the clearer the administrative boundaries become, the easier it can be for problems between those boundaries to belong to no one. Peace is particularly vulnerable to this because success is largely invisible. Institutions work, disagreements are resolved and communities absorb pressure without violence. Nothing dramatic happens. Yet that ordinary stability is produced continually through interactions between people, institutions, rules, expectations and opportunities, many of them affected by policies that would never normally be described as peace policy.
Government has departments. Society has systems. If we want to understand how peaceful societies remain capable of managing conflict without becoming consumed by it, we need to learn to see both.
