Executive-Legislative Deadlocks in the Dominican Republic

Latin American Politics and Society, Summer 2008 by Marsteintredet, Leiv

The baseline problem is to identify the "normal" legislative activity. The overall mean of production of laws for the whole period could be applied as the baseline. This assumes that, all else being equal, different governments of varying colors have the same legislative activity, and that a significant drop in the legislative activity is caused by a deadlock. Using the overall mean as the baseline is problematic, however, because presidents of different colors vary naturally in their activity. Presidents have honeymoon and lame duck periods, political systems experience periods of a limited agenda, and the baseline might depend on the same factors as those that create a deadlock.

The Dominican case presents two specific challenges. First, there is a significant variation in the production of laws between the three parties that have held the presidency since 1978.15 second, Balaguer's administrations were less active than the other administrations. The best way to control for these effects is to use the means of the different parties in government as the baseline for calculating the cutoff points. This approach assumes that different levels of activity can explain the variation between the means of each party in government, and that only the deviations within each of these units can be explained by a deadlock. The problem is, of course, that the low activity is not necessarily voluntary, but might be caused by deadlocks. This measure, nevertheless, alleviates somewhat the problem of not being able to use the box score by controlling for the party in government and the "Balaguer effect."

At the numerical level, it is difficult to argue theoretically for the cutoff point. The cutoff point should establish whether a legislature is deadlocked or not. Given the definition of a deadlock as a serious conflict affecting the output of legislation, the cutoff point should be significantly lower than the normal level of production of laws. One suggestion could be to use a single standard deviation lower than the appropriate mean as the cutoff point. The problem is that the standard deviation is affected by the variable of executive-legislative relations: a high variation of production of laws-that is, potentially a sign of deadlocks-increases the standard deviation. This problem is further aggravated if we use the mean production of laws of the different governments rather than the overall mean as the appropriate mean. Another suggestion could be to use a benchmark number of laws as a cutoff point, but because some administrations pass more laws than others, this is also problematic. Therefore we suggest using a proportion of deviation from the baseline as the cutoff point. If the production of important laws in a legislature is lower than one-third of the baseline, we code the legislature as deadlocked. Whether this cutoff point is adequate can be validated only by a thorough qualitative analysis.

LEGISLATIVE PRODUOTION AND DEADLOCKS IN THE DOMINICAN REPUBLIC

Figure 1 presents the production of important laws and pensions and insignificant laws in the period from 1978 till the end of 2005. Both legislatures of 2003 and the first legislature of 2004 are excluded due to incomplete data. The figure shows that even among the important laws, large variations occur, both within the presidencies and across the different administrations; which indicates that measuring executive-legislative relations as a continuous variable with the mean of production with a presidential term as the time unit might not be the most appropriate for this case.