The analysis of economic loss attributed to the shadow economy has attracted much attention in recent years by both academics and policy makers. Often, multiple indicators multiple causes (MIMIC) models are applied to time series data estimating the size and development of the shadow economy for a particular country. This type of model derives information about the relationship between cause and indicator variables and a latent variable, here the shadow economy, from covariance structures. As most macroeconomic variables do not satisfy stationarity, long run information is lost when employing first differences. Arguably, this shortcoming is rooted in the lack of an appropriate MIMIC model which considers cointegration among variables. This paper develops a MIMIC model which estimates the cointegration equilibrium relationship and the error correction short run dynamics, thereby retaining information for the long run. Using France as our example, we demonstrate that this approach allows researchers to obtain more accurate estimates about the size and development of the shadow economy.