Abstract<\/strong><\/p>\n We study some measurement issues that arise when analyzing the long run behavior of the\u00a0 jth<\/sub> creative region\u2019s time t log output per creative class member (yj<\/sub>(t)) when this region is part of an aggregate economy of j=1,…N creative regions. We focus first (second) on absolute (relative) convergence. In the absolute (relative) convergence case, the N creative regions are similar (dissimilar) in that they all have the same (different) balanced growth path (BGP) level of log output per creative class member denoted by yBGP<\/sup>(yj<\/sub>BGP<\/sup>) In the absolute convergence case, we analyze how to estimate the speed of convergence parameter (\u03c3) and then discuss the relationship between the variance of yj<\/sub>(t) and that of yj<\/sub>(0) In the relative convergence case, we study the error associated with estimating \u03c3 using the methodology of the absolute convergence case. Finally, suppose yj<\/sub>BGP<\/sup>= a + bXj<\/sub> where Xj is an explanatory variable such as creative capital and a and b are positive constants. Here, we study how to estimate b from our knowledge of the coefficients of a related cross-region growth regression.<\/p>\n