Statistical studies using regression analyses have tried to turn correlations into causal stories by measuring putative causes of war at one time and the incidence of war at a later time.

True: published quantitative studies of war before November 2012 (e.g. democratic peace and trade-conflict research by Russett & Oneal, or Hegre et al.) use temporally lagged predictors, measuring the variable at time t and conflict at t+1 or later, to support causal inference. Mostly true: such studies exist, but they use somewhat different designs (e.g. panel lags combined with other methods). Mixed: there are only a few examples, or lagging is used without any causal aim. Mostly false: studies of this kind are rare or marginal, and the claim overstates them. False: no such studies existed.

Set before the research began.

1. E1 Causes of Peace: Democracy, Interdependence, and International Organizations, 1885-1992 ETH Zurich ISN repository (paper presented at the 2001 APSA Annual Meeting, by John R. Oneal and Bruce Russett) · 1 July 2001

   “distributed lags of important liberal and realist variables as well as the dyad’s history of conflict.”

2. E2 Causes of Peace: Democracy, Interdependence, and International Organizations, 1885-1992 ETH Zurich ISN repository (paper presented at the 2001 APSA Annual Meeting, by John R. Oneal and Bruce Russett) · 1 July 2001

   “Granger (1969) proposed that a variable X might be a cause of Y if past values of X can be used to predict Y more accurately than using past values of Y alone.”

3. E3 Causes of Peace: Democracy, Interdependence, and International Organizations, 1885–1992 Oxford Academic, International Studies Quarterly · 21 July 2003

   “Using distributed-lags analyses, we find that economically important trade does have a substantively important effect in reducing dyadic militarized disputes, even with extensive controls for the influence of past conflict.”

Oneal and Russett's 2001 APSA paper, published in International Studies Quarterly in 2003, used distributed lags of liberal and realist variables to explain later conflict (E1, E3). It framed these tests in terms of Granger causality, where past values of X predict Y (E2). From the lagged analyses, the authors concluded that trade reduces later militarized disputes (E3). These are published quantitative studies of war from before November 2012 that use temporally lagged predictors to support causal inference, which is what the criteria require for "true".