656 



TIMBER RESOURCES FOR AMERICA'S FUTURE 



data have thus been subjected to considerable 

 checking and comparison over the years. The 

 data for many earher years are also considered 

 sufficient^ adequate, although the reliability of 

 the consumption estimates probabty varies ap- 

 preciabl}^ from ^'•ear to year, and the later timber 

 cut estimates quoted in the timber section Growth 

 and Utilization are undoubtedly more reliable 

 than the earlier figures. 



Logging residue and plant residue data are 

 also discussed here. Although fewer comparisons 

 or breakdowns can be made with these data, they 

 are believed to be adequate for the purposes for 

 which they are used. 



Data Provided Mainly by the Census 



The Bureau of the Census customarily com- 

 piles output or consumption statistics for lumber, 

 veneer logs and bolts, and pulpwood. However, 

 for 1952, nationwide Census data were not avail- 

 able when needed, and only pulpwood data were 

 obtained solely from this source. Census esti- 

 mates of 1952 Imnber production were used for 

 overall control throughout the West except that 

 Western Pine Association data were used in com- 

 puting saw-log output in part of California. In 

 ttie East, Census statistics were not available in 

 time, and saw-log output data were obtained by 

 other means ranging from 100- percent canvasses 

 of lumber production in some States to adjust- 

 ment of prior year estimates on the basis of tim- 

 ber severance tax reports or limited sawTnill con- 

 tacts in others. For the country as a whole, these 

 procedures resulted in an estimate of 1952 lumber 

 production which is about 5 percent higher than 

 the figure subsequently reported by the Census. 



Output and consumption estimates for other 

 products came from a variety of sources. For 

 example, in the case of veneer logs and bolts, the 

 1952 estimates were derived mainly from Census 

 data, but additional canvasses were made to de- 

 termine the volume of logs and bolts consumed 

 at green veneer and container veneer plants. 

 Estimates of cooperage logs and bolts, poles and 

 piling, round mine timbers, and hewn ties were 

 obtained usually by mail and field canvass of 

 producers or consumers, but, in some instances, 

 severance tax records, public timber sale reports, 

 or ratios based on coal production or wood preser- 

 vation statistics were used. Fuelwood and fence 

 post data came mostlj^ from Forest Survey can- 

 vasses in sample areas. For other products, 

 procedures varied but they were generallj' similar 

 to one of those above. 



For the same product, units of measure varied 

 from place to place, and the basic data applied 

 to different stages in the production process. 

 In order to place all of the statistics for a given 

 product on a standard basis, converting factors 

 were used. These were usually developed product 



by product in special studies made in each region 

 and involving a comparison of local practice with 

 the standard units used in this report. In most 

 regions, such studies are made as a part of the 

 Forest Survey. 



Finally, the timber cut estimates were calcu- 

 lated b}^ adding to timber products output the 

 volume of growing stock that is cut, knocked 

 down, or killed in logging, but otherwise left 

 unused in the woods — the logging residues. The 

 data on logging residues were obtained from 

 Forest Survey studies made on logging operations. 

 Where the Forest Survey had not been made, 

 comparable data from similar logging operations 

 in other States were used. Saw-log and pulpwood 

 logging residues have been studied much more 

 intensively than residues from other kinds of 

 logging. 



Saw- Log Data Have Major Effect 

 ON Sampling Accuracy 



For timber cut in 1952, the sampling accuracy 

 goal was 12 percent per billion cubic feet. Al- 

 though sampling errors could not be computed 

 for all components, the major component is tim- 

 ber products output, which is estimated to have 

 a sampling error averaging 6.5 percent per billion 

 cubic feet. Hence, it seems safe to conclude that 

 the sampling error of timber cut is well under 

 the goal of 12 percent per billion cubic feet, and 

 likely to be 8 to 9 percent. 



The sampling accuracy goal for the estimate of 

 timber products output was set at 10 percent per 

 billion cubic feet. Since it accounted for more 

 than half of the total output of roundwood, the 

 saw-log estimate had a major effect on the achieve- 

 ment of this goal. For all regions and States for 

 which the data provided a basis for computing 

 sampling accuracy, the sampling errors per bilUon 

 cubic feet of output are as follows: 



\'eneer All other 

 Saw Pulp- logs and prod- 

 logs wood bolls uct3 

 (per- (per- (per- (per- 

 Region or State: ccnO cent) cent) cent) 



New England 0. 7 '0 1.0 2. 1 



Middle Atlantic 6.5 1. 4 



Lake States 4.0 7. 6 



South Atlantic 7.4 .8 9.8 



Florida and Georgia com- 

 bined 7.6 1.8 10.0 



California 5.8 .8 (-') 



xMontana 1.1 9. 9 



Idaho .5 17.2 



Weighted average 6. 3 1.1 9.0 



' Zero sampling error indicates 100- percent coverage. 

 - No basis for estimating sampling error. 



In other States and regions, probability sampling 

 was not employed, but the methods and intensity 

 of coverage indicate that estimates of saw-log out- 

 put are well within the sampling accuracy goal of 

 10 percent per billion cubic feet. 



