1. Urban system is very heterogeneous all over the world and so in India. In India, it includes settlements with population size ranging from 5,000 to 15 million. Non-recognition of this and considering urban sector as a monolith sector has led to urban policies swerving towards development issues of metropolitan and large cities. However, small and medium towns are extremely important for balanced as well as rapid economic growth, as sizeable population lives in these towns, and they have good links with the rural economy.

Objectives

2. The study aims at measuring incidence of poverty and changes therein over time, the employment structure and changes in the same and financing of development among the medium and small town in major Indian states. The incidence of poverty in different size class of towns is expected to be related to the socio-economic characteristics of the state, levels of social and economic infrastructure in the states, economic characteristics of the small and medium towns in each of the states and financing of development in these towns which is different for each of the states. The objectives of the study are:

1. To estimate levels of poverty and living in small and medium towns in large states in India.
2. To observe employment characteristics in the small and medium towns in the large states in India and relate the same with level of poverty.
3. To observe the levels of well-being that includes levels of facilities in small and medium towns of the state.
4. To relate levels of social and economic infrastructure in small and medium towns of different states and socio-economic indicators of the states with incidence of poverty in the small and medium towns in these states.
5. For three states in India, to observe the financing pattern in the small and medium towns and relate the same with levels of development and poverty in these towns.
6. To identify areas of intervention by the state government and the NGOs to reduce poverty in the small and medium towns.

Scope of the Study

3. There is general scarcity of analytical studies on the functioning of the small and medium towns. Ideally, a study of these towns at all India level is warranted. This study intends to fill this gap by estimating poverty incidence, structure and nature of employment and analysis of levels of living by linking it to patterns of urban development among seventeen major states of India.

4. The study will cover the period 1987-88 to 1999-00. The effects of liberalisation on the changes in the employment pattern and earnings among these cities and incidence of poverty will also be looked at. Three states, where study of financing of development in small and medium towns will be carried are Gujarat, Andhra Pradesh and Punjab. The scope is limited to only three states for covering this objective is because of the time and resource constraints faced for collecting such information. Such data have to be collected by visiting individual states.

Data Sources and Methodology

For this study, we have used the following classification of urban centres:

1. Small towns – Size class 1 (C1), up to population 50,000
2. Medium towns – Size class 2 (C2), population 50,000 to 200,000
3. Large cities – Size class 3 (C3), population 200,000 to 1 million
4. Metropolitan cities – Size class 4 (CM), population > 1 million

This report deals with four broad aspects of development of small and medium towns. These are:

1. Poverty – Measured through Head Count Ratio (HCR) and Average Per Capita Expenditures of the whole population and of the Poor.
2. Employment – Represented through Work Participation Rates, Unemployment Rates and Nature of Employment
3. Levels of Living – Represented through Basic Facilities such as availability of water supply and sanitation and sewage facilities.
4. Finances and Infrastructure Development – Represented through indicators of finance and basic services.

5. Analysis for the first three aspects is at the state level with figures on all these calculated at the state level from the NSS data. The analysis for the NSS rounds is over a long period of time, more than one decade, starting from 1987-88 to 1999-00, that for the finance and infrastructure data is only for last few years of the decade of 1990s.

6. The analysis of finances and infrastructure development is only for three states in India, Gujarat, Andhra Pradesh and Punjab. In each of these states, the finance and infrastructure data was collected for each urban centre separately and then the analysis for small and medium towns has been done.

7. Considerations for choosing these three states are: (i) In Andhra Pradesh there is an on going intervention for poverty reduction by DFID for class – I towns, some of which are small and medium towns; (ii) Gujarat has attracted a lot of the investment in the post-1991 period and many of the new projects are in rural areas or near urban centres other than the metropolitan cities and (iii) Punjab, which has very low urban poverty, which has well established agricultural marketing networks and has expectedly low disparity between the metropolitan cities and small and medium towns with regards to poverty and basic facilities.

8. For poverty and employment estimates, we have used consumption expenditure and employment & unemployment data of three rounds, 43rd Round (1987-88), 50th Round (1993-94) and 55th Round (1999-00). For levels of living estimation, two rounds of NSS, the 49th Round and 54th Round data are used.

The Main Findings

Level of Poverty (HCR) and Average Per Capita Consumption Expenditure (APCE)

1. The APCE increases with the increase in size class of town. But, the APCE in small and medium towns is higher in the better-off states than the states such as Uttar Pradesh and Orissa.
2. The gap between medium size group of 50,000 to 1 million and metropolitan city has declined particularly in the reforms period (1993-94 to 1999-2000).
3. The disparity of average APCE across state categories measured by coefficient of variation (CV) shows that in 1999-2000 disparity is highest in metropolitan size class. But disparities in other two size classes of towns have risen substantially from 1987-88 to 1999-2000 whereas during this period disparity in metropolitan city size has remained more of less stable.
4. All through the period 1987-88 to 1999-2000, the Head Count Ratio (HCR) in smallest town size of (< 50,000) remains much higher than other higher town size groups and at all-India level, HCR of metropolitan cities continue to be around half of the level of the small towns.
5. Disparity in HCR across states in each of the three town size groups and in all urban areas continues to increase substantially over the years as revealed by CV.
6. In all size class of towns, in all the states except Orissa, HCRs have declined over the three NSS Rounds, during the 12 year period. From 1987-88 to 1999-2000, highest reduction in poverty is achieved in metropolitan city size group. But this reduction is much higher during 1987-88 to 1993-94 rather than during 1993-94 to 1999-2000.
7. In the reforms period, reduction in urban poverty turn out to be much stronger in town size group of less than 50,000 and 50,000 to 1 million compared to the pre-reforms period (1987-88 to 1993-94).

Employment-Unemployment

1. In all size class of towns, male Work Participation Rates (WPRs) are much higher than female WPRs by UPSS (Usual Status), in all the NSS Rounds.
2. The gap between the male and female WPRs is much higher in the metropolitan cities than the small and medium towns. In fact, the gap between the two increases with the increase in size class of towns in all the years.
3. With the increase in size of the town, the WPRs decline among the females. This is true for all the NSS Rounds. Possibly because of male wage rates in the metropolitan cities would be higher than in the small towns, that is male labour is more productive in the metropolitan cities than in the small towns, keeping female members of high wage households in the house.
4. The WPRs have increased for males from 1987-88 to 1993-94, but show marginal decline during 1993-94 to 1999-00.
5. The pattern of increase of WPRs (males) with the increase of size class of town is largely true for states that do not have metropolitan cities. In states, that have metropolises, there is a mixed picture, in some states, in some years, the male WPRs are higher in the metropolitan cities than in small towns and in some states it is the reverse.
6. Among the males, regular employment (per cent RE) increases and casual employment decreases (per cent CL) with the increase in size class of towns. This is true for all the three NSS Rounds.
7. Evidently, more regular employment is available in the larger cities than in the smaller towns to males. Because of the regular employment, income levels would be higher in the larger cities as compared to the smaller towns. This gives an indication as why poverty incidence is higher in the smaller towns as compared to larger cities.
8. In all the size classes, the RE has declined and CL increased from 1987-88 to 1993-94, indicating increasing casualisation of male workers in this period. This trend of casualisation among male workers continues to the end of the decade of 1990s, in small and medium towns whereas in metro cities that gets halted. There is thus casualisation of male labour taking place in the urban economy. But, the increase in proportion of CL in the total workers is much more in the small towns than in the metropolitan cities.
9. In the small towns there is hardly much of RE available to females. In small towns therefore, more than half the females are engaged in SE type of activities.
10. Among females also, RE increases with the increase in size class of towns.
11. Kerala comes out as a state, which has quite high proportion of casual labour among males in each size class of towns and in all the three NSS rounds.
12. The pattern of distribution of workers by industrial categories is quite in tune with our understanding that the metropolitan cities would have the least primary sector workers and the small towns would have the most primary sector workers. The converse is, that the small towns would have the least non-primary workers and metros the most.
13. With the increase in size class of towns, the proportion of secondary and tertiary sector workers increased among the males. But, in the metropolitan cities, it is the tertiary sector workers proportion increases and the secondary sector workers proportion remains the same as in C3 towns. In metropolitan cities, only 1 per cent of the male workers were engaged in primary sector.
14. Among the female workers, the pattern of employment is different. In the small towns, much higher proportion than among the males are employed in the primary sector. Recall that the female workers in these size class of towns were primary engaged as self-employed. This is because they have been employed in the primary sector activities. This shows that when the settlements upgrade to urban status, the males are first to make a transition to non-primary employment and the females continue to work in the primary sector, in agriculture as well as livestock activities.
15. With the increase in size class of towns, female employment in secondary activities increase, but not when the urban centre becomes a metropolis. Then, the female workers tend to get overwhelmingly concentrated in the tertiary sector.
16. Thus, with the increase in size class of towns, proportion workers engaged in manual labour type activities decline. There is increase in managerial type of jobs in the metropolitan cities. This is true for male as well as female workers and is true for all the three NSS Rounds.
17. Thus, with the increase in size class of towns, more skilled and mental labour oriented jobs are available and in the small towns more physical labour oriented jobs are available to the males. The managerial jobs are far higher paid than the manual labour jobs. As a result, there would not only be higher per capita incomes in metropolitan cities as compared to small towns, which also indicate lower poverty levels in the former as compared to the latter.
18. In 1999-00, there is a change. Proportion of male workers engaged in managerial type of jobs increases to 21 per cent from 18 per cent in the previous two rounds, and clerical type of jobs going down to 37 per cent from 40 per cent in the previous two rounds. This is the effect of coming of new highly paid tertiary sector services in the large and metropolitan cities.
19. Improvement in the quality of employment among female workers with the increase in size class of towns is quite marked.
20. There is also overwhelming concentration of females in clerical jobs in the metropolitan cities. It is not the case with smaller size class of urban centres where very low proportions of females have been in clerical type jobs. One can safely say that in the small towns, there is overwhelming concentration of females in manual jobs, mainly primary sector activities whereas in the metropolitan cities, they are overwhelmingly concentrated in the clerical jobs in the tertiary sector.
21. Among the males, in 1987-88 and 1999-00, the UPSS unemployment rates increase with the increase in size class of urban centres.
22. The overall UPSS unemployment rates among males have come down, over this 12 year period in all the size class of towns. In 1993-94, the UPSS unemployment rates are further lower as compared to both, 1987-88 and 1999-00 Rounds.

Level of Well-Being

1. Lowest sized towns were most constrained in this regard in comparison to larger sized towns. Large cities and Metropolises had best in access to tap water supply.
2. But, not necessarily, the metropolitan city households had better quality of access as in several states less proportion of households here than in the small and medium towns had drinking water access in the premises.
3. Developed states had better access to tap drinking water than the under developed states, and small and medium towns of developed states had higher access to tap water than the small and medium towns of under developed states.
4. Adequacy of drinking water available and satisfaction with the drinking water available does not have to do much with the size class of urban centre.
5. Access to bathroom facilities improved from small towns to large cities but not in the metropolitan cities, in which households did not fare better than large cities in the whole period.
6. Access to bathroom facility does not improve with the increase in size class of towns across all the states. This probably would be because availability of bathroom facility is a function of density and metropolises have much higher densities (or crowding), which does not make it possible for each household to have a separate bathroom facility.
7. In regard to drainage facility there is substantial improvements with the increase in size class of towns, the smaller towns having much poorer drainage facilities than metropolises. This would be because, with the increase in size class of towns, the capacity of the Urban Local Bodies (ULBs) to provide drainage facilities increases, especially in the states where the ULBs are powerful. In states where the state level agencies and not the ULBs provide drainage facilities, there also, the per capita expenditure on the drainage facilities in the metropolises tends to be larger than in the small towns.
8. The access to pucca covered or underground drainage is largely limited to metropolises and to a certain extent to large cities. There is substantial improvement in coverage in this regard from tiny small towns to metropolises, once again for the same reason discussed above, of metropolises getting preferential treatment or having higher capability for taking care of drainage facility.
9. Barring Gujarat, in no other states small towns had this facility to considerable extent.
10. Drainage facilities were better in developed states than in the under developed states in all size class of towns.
11. In regard to the access to latrine facilities as well as linkage of latrine to the sewerage system, the data shows clear improvement from small towns to metropolises. Differences among size classes of towns were sharper in case of linkage of latrine to sewerage system.
12. Garbage collection by local authorities is undertaken at a small scale in urban India. Not much difference lies in this regard across size class of towns.
13. At all-India level, access to bathroom facilities improved from little more than half to three-fourth for all households from 1993 to 1998 period. This improvement is observed almost uniformly in all size classes of towns at all-India level.
14. During 1993-98 period, there is substantial increase in drainage facilities in small and medium towns, that has reduced somewhat the disparity between the metropolitan size class and small and medium towns.
15. Improvement in latrine facilities from 1993 to 1998 is uniform across all size class of towns.

Municipal Finances

1. Size of municipal budgets increase substantially with the increase in size class of urban centres. But, more important than this is the fact that the per capita budget size increases with the increase in size class of urban centre. This is very much true for Gujarat and Punjab, but not fully true for Andhra Pradesh, where, because of the state government grants to the small and medium towns, the per capita budget size in small towns is larger than the medium towns.
2. In Andhra Pradesh, the average budget size in towns below population of 50,000 is larger than the same in Gujarat and Punjab. But, this is not true for higher size class of urban centres.
3. Gujarat had the largest per capita budget size in all size classes except the small towns.
4. In Punjab, in all size classes, there is very little fluctuation in the division of budget into capital and revenue account shares. In Gujarat and Andhra Pradesh, there is a very high fluctuation in the share of the capital account in the total budget. This is true for all size classes of urban centres.
5. On the whole, across all the size class of towns, the total budget size has increased in Punjab in real terms. In Gujarat, this is true only for the small and medium towns and not for the metropolitan cities.
6. The level of municipal income is thus an important determinant of the levels of services provided by them. To understand the local bodies’ ability to provide services, it is necessary to study the trends in municipal income of the different size of towns. The percentage share of own sources revenue is quite high in municipalities of all classes in Gujarat. With the increase in size of the ULB, the share of own source revenue in the total income increases. This is true for all the years of our study.
7. In Punjab, almost whole of income of the local bodies are from internal sources; very negligible amount (and even nil in some classes in some years) of the total income came from external sources. Even the smaller towns have very small share of their income coming from external sources. Thus municipalities in Punjab, across all classes have very less dependence on external sources for income.
8. In Andhra Pradesh, all size class of ULBs have much higher dependence on the external sources of revenue as compared to other two states. Further, the smaller towns here are more dependent on external sources of funds than the bigger towns.
9. In general, tax income is expected to be the main source of own source revenues of the ULBs. And that indeed is the case.
10. Sources of external income for a local body are grants and contributions from higher levels of government and loans from government or other agencies. In Gujarat, the per capita external income is higher for larger urban centres than for the smaller towns.
11. Per capita grants is highest in case of Andhra Pradesh, followed by Gujarat and finally Punjab.
12. State government contribution to local bodies in Gujarat are highest in case of the fourth size class, which means that the local bodies having population more than 1 million are benefiting more from grants.
13. The amount of State Government contribution is very less in Punjab, and it is absent in case of the last two size classes. It is only the small towns which gets some grants and contributions and that amount is also very less.
14. Per capita State Government contribution to local bodies in Andhra Pradesh are quite high, and it is highest in case of the smallest towns.
15. Per capita loan income is the least in case of Andhra Pradesh, followed by Punjab.
16. Local bodies in Gujarat have comparatively higher loan incomes, especially in the case of the local bodies in the last size class.
17. In Gujarat, even small towns have taken loans.
18. Per Capita expenditures, total, and especially revenue increase with the increase in size class of towns.
19. With the increase in size class of urban centre, there is increase in per capita expenditure on general services which includes administrative costs.

State Level Correlations and Policy Implications

9. The above analysis shows that in all regards that is with respect to various dimensions of development there is significant disparity across different size class of urban centres. In fact, level of development in urban centres is very much a function of the size of the urban centre. This is true for all the states in India, where the small and medium towns are at a greater disadvantage as compared to the large cities and particularly with regards to the metropolitan cities. And the gap between the metropolitan cities and the small towns is quite large with regards to all aspects of development.

10. This brings us to a crucial conclusion. All urban policies, whether on urban employment, on urban poverty alleviation, on urban facilities and on financing of urban facilities cannot be the same for all sizes of urban areas. Urban sector has to be viewed at a disaggregated level. Till now, due to lack of such disaggregated data, scholars and policy makers based their analysis of urban areas only at the state level, relating urban sector performance with the state’s development process or looked at some urban indicators at the district level. Now we have new set of data that gives levels of development in different size classes of towns. And this makes disaggregated policy making possible.

11. We have analysed what are the critical factors that would lead to reduction of poverty in the small and medium towns. We have done this by simply calculating the correlations between the HCRs for the different size class of towns with five sets of variables; namely (i) per capita consumption expenditure (ii) employment nature and type, (iii) education levels, (iv)other indicators of economic development and composition of economic development and (v) urbanisation and urban settlement structure. This analysis is at the state level and it indicates what type of development at the state level has potential to reduce incidence of poverty in the small and medium towns in the respective states.

12. Employment structure within the size class significantly influences incidence of poverty (HCR) within that size class. In small and medium towns, that is for towns upto population of 200,000, nature of employment variables, especially of the females, have shown a significant relationship with HCR of their respective size class. That is, in towns upto 50,000 population and towns with population between 50,000 and 200,000, HCR and percentage employed as casual labour (females) have significant positive relationship. In contrast, in these two size classes of towns, percentage females employed as regular workers have significant negative correlation with their respective HCR in 1993-94 and 1999-00. An economic development process that ensures regular employment of females in small and medium towns therefore would go a long way in reducing incidence of poverty in these towns. Percentage male workers engaged as casual workers in small and medium towns show a significant (positive) correlation only in 1987-88.

13. As far as sectoral distribution of workers is considered then, that of male workers show a significant relationship with HCR. Percentage male workers in primary sector show significant positive correlation with the HCR in 1987-88 and 1993-94. In 1993-94, even percentage female workers engaged in primary sector show significant positive relationship with HCR. In contrast, percentage male and female workers engaged in tertiary sector in 1987-88 and 1993-94 and only female workers in 1999-00 show significant negative correlation with the HCR. In essence, if workers and particularly female workers get regular employment in the tertiary sector in the small and medium towns will lead to reduction of poverty in these towns. What exists today is that there is very large proportion of female workers engaged in primary sector in the small and medium towns, which has resulted in continuous high levels of HCR in these size classes of towns.

14. As far as quality of employment is concerned, it is interesting to observe that among the males, if percentage employed as manual labour is high then HCR is high in the small and medium towns. This correlation was found to be significant in 1987-88 and 1993-94. In 1987-88, this was true also for females. In contrast, increase in percentage workers employed in managerial jobs among females in the small and medium towns, will lead to reduction in incidence of poverty in these towns.

15. Education variables that show clear relationship with poverty is percentage population with graduate and above level of education, which is negatively and significantly correlated with HCR in small and medium towns. Hence, improvement in human capabilities in the small and medium towns would attract high quality of employment in these towns. In metropolitan cities, we do not find any significant correlation of education variables with the HCR.

16. Economic development variables do not show any significant relationship with the HCRs of the small and medium towns except the variable of tractors per lakh population, an indicator of industrialisation and prosperity in agriculture, which is negatively correlated to HCR of the towns with population less than 50,000 in years 1987-88, 1993-94 and 1999-00. It can be argued that agriculture prosperity might spill over to small towns, thereby leading to decline in incidence of poverty in them. The Human Development Measure developed by Hirway and Mahadevia (2003) for the year 2001, has significant negative correlation with the HCR of C1 (< 50,000 population) towns. This, once again emphasises the point of improving human capabilities in small and medium towns to reduce incidence of poverty.

17. Lastly, we have observed the correlations of urban settlement pattern and urbanisation level with the HCRs of small and medium towns and HCRs of metropolitan cities. This analysis also gives us an entirely different set of variables influencing incidence of poverty in the two. Spread of trading activity and increase in number of settlements whose economy is predominantly or solely trading activity based and increase in population living in such settlements leads to reduction in poverty in small and medium towns. This means that the economic base of small and medium towns can be largely trading and regular, high-quality employment has to be generated in these to reduce poverty in small and medium towns. In contrast, if there are large numbers of settlements with agriculture as predominant or sole activity and there is large number of population residing in them then there is high incidence of poverty in small and medium towns. This is what has been observed with regards to variables of employment.

18. The conclusion of the above analysis is that reduction in poverty in small and medium towns cannot be left to dynamics of economic growth alone. Policy interventions in human capabilities development and employment sector for ensuring good quality regular employment in the tertiary sector in these towns is necessary to reduce poverty in them.

City Level Correlations and Policy Implications

19. The correlations analysis shows that in Gujarat, the population size of towns is related to per capita finance data. In other words, financial position of the towns, especially the income of the town increases with the increase in its population size. Both, the own source revenues as well as grants (on per capita basis) increase with the increase in size class of towns. Also, the per capita loan availability also increases with the increase in size. Increase in per capita income leads to increase in per capita expenditure. And, all these reflect in improvement in the level of at least one facility variable, per cent population covered by drainage facility, with the increase in size of town. Thus, there is an inherent bias in the urban settlement system and urbanisation pattern, the system favouring the large cities over small and medium towns in the state.

20. As compared to Gujarat, in Punjab, lesser numbers of variables of finance were related to the population size of the towns. Thus, all size classes of towns were able to generate their own resources and expend them for various town level functions. But, the towns with higher population growth rates were related to the variables of finance, indicating that the towns having better financial position were in a position to attract new population. Finally, like Gujarat, larger towns had better sewerage facility as compared to smaller size because of their strong links with indicators of city level income. Very unclear picture emerges from the correlation analysis of different set of variables in Andhra Pradesh. While the finance variables are related to population size, just like Gujarat, and in spite of the class averages showing a different picture, the facilities variables are not related to either population size or to the variables of finance. It appears that the state government is taking care of these functions. Thus, a study of municipal finances in Andhra Pradesh is not of much consequence.

21. Gujarat portrays a picture of the way a market would function. The large cities, having better economic base than the small cities, would be able to attract more incomes and therefore would have higher expenditures and thus better level of facilities. But, worse still, the state policy follows the market discrimination against the weak, and the state government has been giving higher per capita grant to the larger cities as compared to the small towns. The large cities have been also able to get higher per capita loans.

22. Andhra Pradesh portrays the case of achievements through state policy and intervention, wherein urban centres with weaker economic and therefore financial base can have improved level of facilities. But, it also suggests that the urban development only through state government support may not take the urban centres too far. Hence, though the per capita incomes in this state are high in the small and medium towns, especially in the small towns, they are not efficiently expended with quite high expenditures incurred on administrative purposes.

23. Punjab portrays a picture of balanced market-led development, where, first of all, the urbanisation structure is equitable. The small and medium towns in the state are linked to agricultural economy and have quite high per capita incomes, generated through their own efforts. The expenditures in the towns are also therefore on services that would improve the well being of the local population and not much on administrative purposes. The small and medium towns here therefore have high level of services.

24. In short, the financing and level of facilities in the three states studied here reflect the development philosophy of each of the three states. Gujarat’s development has been highly inequitous and this gets reflected in the discrimination against the SMTs. Punjab’s development is based on ‘agriculture-first’ theory where there is a link between rural and urban economies and the urban economies are also based on small-scale industries, both of which bringing in fair amount of equity in the urban settlement pattern and level of development across different size classes of urban centres. Development in Andhra Pradesh is State-led, the government not only taking the responsibility of urban development but also having strong welfare policies, both of which benefiting the SMTs.

Overall Policy Implications

25. This study strongly brings out the case for disaggregated policy making for the urban sector. This is particularly the case when in this period of economic reforms, when the urban development and poverty alleviation are thought to be taken up through initiatives at the local level, particularly of the urban local bodies themselves, it becomes necessary that the capabilities of each size class of urban centre is understood well. This will lead to a more involved and disaggregated policy making with regards to urban areas. As such, in India, we do not have yet a fully announced urban policy. Housing policy drafts have been prepared but not yet formalized. This study will therefore be of great relevance when such an exercise is taken up. Even in the area of urban poverty and urban employment, there are policies applicable to the entire urban sector, which as this study indicates is highly unequal. Hence, in these two areas as well, separate policies for large and metropolitan cities and small and medium towns are required.

26. Specific policy for the SMTs is expected to be the IDSMT scheme. As the review of the IDSMT scheme suggests, not much of coverage exists in this scheme. This study points to the fact that policy action is required on many different fronts to address the issue of development of SMTs. The study points at holistic view and policy making with regards to the SMTs.

27. Lastly, this study also brings out the fact that there is a need to streamline the definition of urban, as population census and urban development department of the state government have different definition of urban. Urban policies should be applicable to only settlements that have urban local bodies and not all settlements that are declared as urban by the population census. Hence, this study addresses multiple levels of policy issues, one with regards to definition of urban, with regards to data sources for urban research and with regards to various dimensions of development in especially small and medium towns.