While examples of regional trade barriers can be found within Russia, the pervasiveness and persistence of such barriers is unclear. We investigate this issue by analyzing the extent to which regional borders and distance account for observed differences in commodity prices across 74 regions in Russia. The results indicate substantial variation across time: an initial period of widespread integration gradually gave way to a period of disconnectedness during 1995 through 1997, which had apparently subsided by May 1998.

Next we analyze how these temporal fluctuations in market integration relate to temporal fluctuations in inflation volatility, internal transport costs, international trade, public discontent, and standards of living. With the exception of inflation volatility, we find that these variables exhibit strong statistical relationships with integration. Regarding the relationship between integration and international trade, the widespread disconnectedness observed between 1996 and 1997 coincides with a period when Russia exhibited strong export performance and attracted a large amount of international portfolio investment. A possible interpretation is that the unusually high trade to GDP ratio could signal weakening inter-regional trade relations in Russia.

A growing literature has attempted to measure and explain market integration in transition economies using a variety of different approaches. Several studies have noted a high degree of inter-regional price dispersion, leading the authors to be pessimistic about the opportunities for market integration in Russia. Our goal in studying temporal changes in market integration is to begin to understand the causes of these changes in Russia.

In integrated market economies, arbitrageurs and traders can quickly move tradable goods from regions in which sales prices are low to regions in which prices are high, so long as transport costs are not prohibitive. This implies that within a group of integrated regional markets, price differentials for similar tradable goods sold in different regions will increase in proportion to the distance separating the regions.

Our goal is to measure temporal movements in inter-regional price dispersion. We pursue this goal using a regional dataset that consists of two commodity price indexes: an index that measures the cost of a uniform basket of basic food goods and a regional consumer price index that includes a basket of food goods, nonfood goods, and services. The food basket data span the period April 1994 through November 1999; the regional consumer price index data span the period January 1994 through November 1999. The dataset includes observations collected monthly from 74 Russian cities: Moscow, St. Petersburg, and capital cities in 72 other regions.

The more comprehensive picture of integration we obtain is, as before, roughly U-shaped: an initially high level of integration existed until June 1995, it then fell markedly and remained relatively low until approximately June 1997, and then climbed fairly steadily through the remainder of the sample period. An interesting finding was that the financial crisis of August 1998 had no noticeable impact on integration dynamics.

Our interest in relation to market integration and the aggregate economy is in analyzing how the temporal fluctuations in regional market integration correspond to related economic and political variables. In earlier studies we showed that voting patterns in the June 1996 presidential elections were important in explaining the regional isolation observed between 1994 and 1996.

Specifically, we classified Russian regions into two categories: those that supported the Communist Party candidate Zyuganov in the final round of elections (termed Red Belt regions), and those that supported Yeltsin. Zyuganov's platform emphasized price controls, price subsidies, and administered resource allocation. In contrast, Yeltsin's platform called for further price liberalization, a cutback in price subsidies, and a deepening of the privatization process. We found that this difference in platforms was an important predictor of differences in policies between the Red Belt and the rest of Russia.

Specifically, we observed significantly more retail price regulation, budgetary and agricultural price subsidization, and less entrepreneurial activity (as measured by the number of legally registered small private firms per capita) in the Red Belt during 1995-96. A possible explanation of this pattern might be that regional governments that use subsidies and price controls to keep basic consumer goods cheap have an incentive to erect borders to limit the outflow of their cheap goods via exports or consumption by nonresidents. The presence of price restrictions and border controls is likely to result in an economic environment that is not conducive to entrepreneurial activity.

To expand on our earlier findings, we begin our analysis here by studying whether regional reformist voting patterns correspond with the evolving pattern of integration reported above. In the June 1994 subperiod, 57 percent of the proreformist voting regions were integrated according to our measure, while 43 percent of the nonreformist regions were integrated.

In the June 1995 subperiod the integration percentages were 24 and 8 percent, respectively. We also find a significant difference in integration percentages for the June 1998 subperiod (66 versus 46 percent). However, this is not the case for the remaining subperiods: integration percentages were nearly identical across classifications in the June 1996 and 1997 subperiods, and in the May 1999 subperiod the integration percentage was actually lower in the proreformist regions (73 versus 86 percent). Thus our findings regarding integration and political attitudes toward economic reform are mixed: the clear positive correspondence observed early in the sample does not hold in the later part of the sample.