Calculating the consequences of global public bads such as climate change or pandemics helps uncover the scale, distribution and structure of their economic burdens. As violent conflict affects billions of people worldwide, whether directly or indirectly, this article sets out to estimate its global macro-economic repercussions. Using a novel methodology that accounts for multiple dimensions of war, the article finds that, in the absence of violent conflict since 1970, the level of global GDP in 2014 would have been, on average, 12% higher. When disaggregating these results by conflict type, civil conflicts are estimated to have been the costliest by far. Income growth is found to be altered up to four years following the end of a conflict, although the direction of this relationship depends on the intensity and type of conflict. Countries also suffer significantly from fighting in neighbouring countries, thereby showing the importance of mitigating spillovers rapidly. The largest absolute losses associated with violence emanate from Asia, while many high-income economies are found to benefit economically from participating in conflicts on foreign soil. This analysis thus shows that, despite some evidence of a faster post-conflict growth and possible benefits for external participants, violent conflict leads to net global losses that linger long after peace is achieved, reducing the peace dividend. The article concludes by discussing public policy options to strengthen the benefits of peace as a global public good.