Pages:     || 2 | 3 | 4 | 5 |   ...   | 9 |

Chapter 5. One alternative approach to typology of the RF regions

One of the major problems with geographically large countries such as Russia and Canada is regional inequalities. This results in major problems of socio-economic-political integration. The traditional friction of space is a tremendous barrier to spreading wealth equally throughout a nation. The bigger the country, the greater are its inequalities. A byproduct is that unique regions with particular strength can develop within such nations. To tap this strength, unique regional development policies may have to be developed that are not national in nature. But before this can be done a thorough understanding of the spatial differences in terms of weakness and strength needs to be achieved. A detailed typology study is the required first step. In the following section we wish to present some different models for creating a typology of the regions of the Russian Federation using a basic principal components/cluster analysis approach.

As stated above large countries tend to have greater variation between the various sections of their domains than smaller countries. Spatial variation in socio-economic well being increases with the size of a country. In part, it is the outcome of a spatial law which states that nearer things are more related then further things. It is, in a sense, an outcome of the law of gravity, which states that nearer things attract each other more then further things. This seemingly holds true in the human sphere of influence as well as in the physical world. Hence, richer people live near each other, businesses tend to concentrate, and poor regions are usually found in proximity to each other. Spatial autocorrelation occurs in nearly all variables that spread over space.

One of the main tasks of a federal government is to provide equal opportunities for all its citizens, no matter where they are located. Thus, it is normal that it will attempt to relieve spatial variation among the well being of its citizens by means of various forms of regional development and/or social transfer programs. On the other hand, local or regional governments see it as their mandate to give their citizens the best standard of living possible irrespective of the conditions in the other regions. As a consequence of the variations of natural and man made endowment factors over space, inequalities will evolve between regions.

The reasons why some regions in the world are developed socio-economically while others are not, are still heatedly debated among academics. Theories, models and concepts abound that try to explain regional economic development differences. Some of the more important ones are: Growth Pole (Perrou, 1950), Competition (Smith, 1776), Circular and Cumulative Causation (Myrdal, 1957), Comparative and Competitive Advantage (Porter, 1990), Core-Periphery (Friedmann, 1966), Economic Base (Richardson, 1973), Growth Stage (Rostow, 1960), Entrepreneurship (Schumpeter 1944), Trade (Ohlin, 1933), Backward and Forward Linkage (Hirshman, 1958), Staple Growth (Innis, 1930) and Central Place (Christaller, 1933). A number of these have modern derivatives as well.

These theories/models/concepts of how growth and development can take place are by no means all the ones that could be listed. In addition, they are not separated into pure growth and development models. However, among Canadian economic geographers and regional economists Staple Growth Theory is the most prevalent model used when trying to explain development in Canada in a historic and regional setting. However, it seemingly has lost its power to account for our present growth patterns and the resulting inequalities in Canada. Since Russia is also a large country with many natural resources, it is tempting to try to build regional development policies on it. For this reason, a short review of Canada’s experience may be useful.

What then is the essence of Staple Growth development theory Harold A. Innis, an historical economist, first proposed the concept in his book entitled The Fur trade in Canada: An introduction to Canadian Economic History, published in 1930. In it he argues that Canada was explored because of the demand for furs in Europe. The money that furs brought in was used to create a ‘civilized’ Canadian society. The latter referred to a way of life that was equivalent to that in Europe and the USA. It cumulated in the ability of central Canada to build the CPR railway across Canada by 1885, thus forming and binding a nation together. Innis argues this elegantly in his earlier book The History of the Canadian Pacific Railroad, first published in 1923. Then the concept was expanded by others to include other natural resources that where exploited and sold abroad. Finally, it was proposed that the export of resources became the staple growth medium for the Canadian economy, hence the terminology of a Staple Growth theory for Canadian development.

Which have been the resources staples that fed Canadian Economic growth Clearly they are; fish, furs, lumber, wheat, forest products, minerals, and of late, energy. It has been suggested that Canada has followed this somewhat unique path to its present day high standard of living, a pattern shared, but only in part, by Australia and New Zealand. The staple growth theory suggests that Canada became wealthy through the sequential sale of its abundant natural resources/staples. In fact, many people around the world still associate Canada with the extraction and export of natural resources. This image has been so strong that the Canadian economy has often been described as consisting of ‘hewers of wood and carriers of water’.

In comparison to other members of the G8 countries, the 'resource image' is probably still somewhat true. However, in Canada, the importance of natural resources is rapidly declining in importance as a share in employment and GDP when compared to other sectors of the economy. At present, total direct employment in the resource sectors (agriculture, forestry, mining, energy, fishing, hunting, etc) contribute no more than 8% of total employment in Canada. It has especially declined in importance during the post-industrial era when manufacturing became less important in the total economies of developed nations.

Therefore, in the future, regional development will have to rely far more on what Porter (1990) calls competitive advantages of communities. It now involves strong human, institutional, environmental, economic and historical development factors for a region to develop. These factors all tend to have systematic regional patterns over space. In contrast to physical or natural environmental advantages, which could not be changed by human hands, these can. Competitive advantages of one region over another can be and have been created in the past.

In order to determine a region’s competitive advantage or disadvantage, one needs to examine its total infrastructure in comparison to other regions. Since human and business factors are of great importance in a region’s competitive mix, any analysis of the competitive nature of regions needs to have a large number of socio-economic variables available for analysis that describe the regions. Even though Canada has a long history of regional development policies and regional payment transfer system, inequality has not been removed. At best these measures have prevented the conditions worsening. Each province in Canada tries to extract as many resources from the federal government as possible in order to increase the well being of its citizens. But the federal government, through agreements with the provinces and through unilateral decisions, regulates these flows of funds. Nevertheless, equality of opportunities for all citizens, no matter where they live, is the underlying principle. Such principles relate mainly to health care, child and unemployment support, welfare, pensions, and access to various federal government services. Presently, the federal government’s regional development policies and efforts are administered through four regional crown corporations, one in Atlantic Canada, one in Quebec, one in western Canada and one, FEDNOR, in northern Ontario.

In 1989 Hecht and Boots published a typology study of Canada in which we attempted to determine which regional development forces were the more important, the federal government’s attempt at trying to make things similar over space for all Canadians or the provincial aims of making each province different from the rest. If the former were stronger, one would anticipate that variation of socio-economic conditions over space would display a spatially random pattern. On the other hand, if the provinces were building unique conditions for their people, the conditions should exhibit clustered patterns in which the spatial clusters would correspond to provincial territories.

To test for these hypothesis they collected 25 socio-economic rate variables for 260 census divisions encompassing all provinces using 1981 Statistics Canada data. Rate variables were chosen to counter the large variation in the population sizes of the census divisions. The variables represented six broad categories; employment, economic, demographic, housing, cultural, and education.

A discriminant analysis of the data brought out four canonical functions with eigenvalues greater than one, explaining 94.1% of the total variation. When the census divisions were grouped, they corresponded strongly with provincial territories. Only a few census divisions grouped with those of other provinces. This led us to conclude that broad province-building forces are extremely strong (Boots and Hecht 1989, 194). A further analysis brought out five major regions in Canada; the Atlantic region, Quebec, Ontario, the Prairies, and British Columbia. The similarity of the census divisions within these regions is substantially greater then between the regions. Again, only a few census divisions were>

Given the Canadian experience, it will be interesting to see if the 89 Russian regions show similar spatial cohesion in characteristics and conditions. Having had a spatially planned economy until recently should speak for less variation over geographic space. On the other hand, the huge size of the country, with its great physical diversity, its cultural mosaic, and its development under a market economy over the last 10 years should produce increasing variations over time.


The data relate to 89 regions. Values are recorded for various years from 1985 to 2000 for 48 variables. However, for some of the years, information is missing for some regions on a number of variables. In view of this situation, we created three smaller data sets for exploratory analysis. The first (Russia99M1) was composed of 88 regions and 24 variables for 1999. This is the most recent date for which extensive information was available. The 24 variables are listed in Table 5.1. The region omitted from this data set was Chechnya. This is because no information is provided for this region for 20 of the 48 variables. Further, for five of the 28 variables for which information is available, the value for the region is a statistical outlier.

The other two data sets consisted of 87 regions and 14 variables (see Table 5.1). The 14 variables are a subset of the 24 variables in Russia99M1 and were selected because they were available for two dates, 1995 (RussiaSmall95) and 1999 (RussiaSmall99). Ideally, we would have liked to examine data for 1992 since this was the first year after the shift from a planned to a market economy. However, only eight variables were available for all regions for this date. The first date after 1992 for which a reasonable number of variables was available was 1995 and so this year was selected. The second date was chosen for the same reason as the larger data set described above. The two omitted regions were Chechnya and Dagestan. The latter was omitted because in 1995 its situation in terms of missing variables was similar to that of Chechnya.

Each data set was analyzed using a two-step procedure. First, a principal component analysis with varimax rotation was applied to the variables. Regression factor scores were computed for all components with eigenvalues greater than one. Then, using Ward’s hierarchical clustering procedure, the regions were grouped into>



Factor Analysis

Principal component analysis of this data set results in six factors with eigenvalues greater than one. Collectively, these six factors explain 80.4 per cent of the variance in the 24 original variables (see Table 5.2). The composition of these factors is shown in Table 5.3.

The first factor, which explains 26.0% of the variance, is dominated by four variables AVLSUB, RETCAP, AVEINC, and OTHERINC. It may be interpreted as a wealth and consumption dimension. Outlier values for this factor are all positive and occur for Moscow, Yamal Nenetsk AO, Khanty-Manslysk AO, Tyumen oblast, and Samara oblast.

The second factor (24.6% explained variance), with high positive loadings of DEMLOD, ELDABA, MIGINC and SOCTRS, and high negative loadings of SUBLEV and WAGSAL, identifies regions with relatively older population dependent on social transfers. One may label this as a human and economic dependence dimension. Outlier factor scores are all negative, implying an absence of the conditions summarized by this factor, and occur for Chukotsk AO, Magadan oblast, and Yamal Nenetsk AO.

Factor three (13.7% explained variance) has high positive loadings for YOUABA, ENTLOS, REGUNE, UNERAT and POPSUB and a high negative loading for ACCPER. It identifies regions of high unemployment associated with higher proportions of enterprises with losses. These regions also have higher proportions of their populations with incomes below the subsistence level, higher proportions of children, and low housing space per person. One could label this dimension as an impoverished employment/housing dimension. Outlier values for this factor are all positive and occur for Ingushetia, Aginsk Buryat AO, Tyva, Dagestan, Ust-Ordynsk AO, Gorny-Altai, Kalmykia, and Taymyr AO.

Pages:     || 2 | 3 | 4 | 5 |   ...   | 9 |

2011 www.dissers.ru -

, .
, , , , 1-2 .