Measuring Nineteenth Century Regional Economic Growth: Historiography and Methods

Abstract

Introduction. The scope of regional economic inequality, its causes and consequences are relevant issues in the economic history. High regional inequality impedes representative estimation of national economic development and international comparison. The end of 19th and beginning of 20th centuries was the time when industrialization, states’ economic and political integration led to their regional divergence/convergence. Methods. The main challenge of measuring and accounting for 19th century regional economic growth is a scarcity of regional historical and economic statistics. Thus, the paper concerns with historiographical analysis of successful attempts to face this challenge in economic history. Results. It can be distinguished three approaches to historical regional economies accounting depending of relevant statistics availability: 1) for countries with high regional-data integrity, GRP can be estimated as a sum of its residents’ incomes (R. Easterling’s method); 2) for countries with moderate regional statistics being saved, it is possible to estimate GRP through distributing known GDP totals across regions on the basis of indicators of regional sectors’ shares (Geary-Stark method); 3) for countries with poor regional historical statistics it fits only short-cut approach on the basis of indirect regional economic indicators (Crafts’ approach and Good–Ma method). Furthermore, the paper deals with following methods and models used in quantitative explorations of unequal regional economic development: shift-share analysis, β and σ-convergence. Discussion. It appears that historical statistics from the Governors reports makes possible to distribute known national values added in the first and secondary sectors across provinces of the late-nineteenth century Russian Empire in the line with Geary–Stark methodology. The contribution of tertiary sector to the provinces’ economic growth could be estimated on the basis of indirect indicators from the same historical source and the other sources, following Good–Ma methodology. Finally, the cross-checking of the GRP to be calculated is possible through comparison with A. Markevich estimates for 1897.