firms be possible, the manufacturing and non-manufacturing payrolls were 

 totalled and the results treated as an index. The average payroll per sample 

 is the figure itself, but Table 9 is only a set of relative numbers. The num- 

 bers are proportional to the payroll data. Trend data were not available. 



The results are in conformity with the hypothesis that the population 

 balance will be independent of the economic base of the town. In 1950 

 the average number of both manufacturing and non-manufacturing es- 

 tablishments in the sample of increasing towns was 2.6; in the decreasing 

 towns, 4.5. The average annual payroll in the sample of increasing towns 

 is $118,250; in the decreasing towns, $152,500. However, since this pay- 

 roll in the decreasing towns apparently stems from seasonal recreational 

 employment, it undoubtedly represents payments to a substantial number 

 of migrant or temporary workers. While the evidence is not conclusive, 

 there appears to be some indication that the economic bases of the two 

 samples are not identical, although there may be considerable overlap in 

 the nature of their economic activity. The data on employment status in 

 1950 tends to support the belief that the two sets of towns have approached 

 some sort of equilibrium which is reflected in the population characteristics. 



These data indicate no significant difference in the percent employed, un- 

 employed, and not in the labor force for either sex in the two samples. The 

 slight differences which do exist are in favor of the sample of increasing 

 towns. 



Farm Size and Value 



Since these are small towns in New Hampshire, it was assumed that agri- 

 culture would play an important economic role, and an important non- 

 demographic variable was that of farm size and value. The period 1920- 

 1950 has been characterized by a dollar whose value has fluctuated con- 

 siderably and so all comparisons in which value of property plays a part 

 have been made with dollars standardized with the General Wholesale Price 

 Index, all commodities, 1926 =r 100. 



Since 1920 the average size of the farms in acres in the sample of in- 

 creasing towns has decreased slightly although the trend is one of fluctu- 

 ation rather than regular decrease. During the same period the average 

 size of the farms in the sample of decreasing towns has increased with a 

 fluctuating trend also. See Table 11. 



The relationship between average number of acres has varied consider- 

 ably. When the significance of the differences between these numbers is 

 analyzed, however, the change in farm size does have a definite trend 

 toward larger farms in the decreasing towns and small farms in the in- 

 creasing tovv'ns. 



Farm Numbers 



The trend in farm numbers in both the increasing and decreasing towns was 

 downward. However, the rates of decrease were different for the two samples. 

 The average percent changes between 1920-1945 in number of farms in 

 the increasing towns was — 14.9, while in the decreasing towns it was 

 — 41.5. When the significance of the difference between average numbers 

 of farms in the increasing and decreasing towns is computed for each period 

 1920-1945, the trend indicates that the difference in farm numbers is de- 

 creasing. See Table 12. 



Obviously the trend over time is for the differences in numbers of farms 

 to approach some sort of minimum. It would be expected in line with na- 



16 



