General Considerations Relating to Secular Trend, 1903-18 Source: The Review of Economics and Statistics, Vol. 1, No. 1 (Jan., 1919), pp. 38-39 Published by: The MIT Press Stable URL: http://www.jstor.org/stable/1928761 . Accessed: 15/05/2014 21:25 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to The Review of Economics and Statistics. http://www.jstor.org This content downloaded from 194.29.185.26 on Thu, 15 May 2014 21:25:44 PM All use subject to JSTOR Terms and Conditions
Transcript
General Considerations Relating to Secular Trend, 1903-18Source: The Review of Economics and Statistics, Vol. 1, No. 1 (Jan., 1919), pp. 38-39Published by: The MIT PressStable URL: http://www.jstor.org/stable/1928761 .
Accessed: 15/05/2014 21:25
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp
.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].
.
The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to The Review ofEconomics and Statistics.
http://www.jstor.org
This content downloaded from 194.29.185.26 on Thu, 15 May 2014 21:25:44 PMAll use subject to JSTOR Terms and Conditions
38 THE REVIEW OF ECONOMIC STATISTICS 38 THE REVIEW OF ECONOMIC STATISTICS
TABLE H -CONSTANTS IN THE EQUATIONS OF THE LINES OF SECULAR TREND, I879-96 AND I897-I913, AND STANDARD DEVIATIONS OF CYCLES FOR SEVENTEEN ANNUAL SERIES.
("CYCLES" MEAN DIFFERENCES OF ORIGINAL ITEMS FROM CORRESPONDING ORDINATES OF LINEAR SECULAR TREND)
y =it + b Standard deviation
SERIES * UNIT |879-I896, Origin I879 I897-1913, Origin I897 of cycles
m b b i879-i896 T897-I9M3
I. General Wholesale Prices ........... Relative to I890-99 -2.I0 I317 + 2.60 96.3 5.20 3.65 2. Gross Receipts of Railroads ........ $Iooooooo +3.6I 6I.i +I2.4 II2. 6.47 8.25
3. Net Earnings of Railroads ......... $IOOOOOOO + .6i6 25.2 + 3.23 39.7 2.o8 6.46 4. Exports of Mdse.. $iOooo0ooo + .846 74.7 + 7.2I III. 7.8I I3.5
I. General Wholesale Prices ........... ........ .48 2.9 .... .... I2.7
i6. Indices of Crop Yield per Acre.+ .432 95-3 .... | *** 7.8
TABLE H -CONSTANTS IN THE EQUATIONS OF THE LINES OF SECULAR TREND, I879-96 AND I897-I913, AND STANDARD DEVIATIONS OF CYCLES FOR SEVENTEEN ANNUAL SERIES.
("CYCLES" MEAN DIFFERENCES OF ORIGINAL ITEMS FROM CORRESPONDING ORDINATES OF LINEAR SECULAR TREND)
y =it + b Standard deviation
SERIES * UNIT |879-I896, Origin I879 I897-1913, Origin I897 of cycles
m b b i879-i896 T897-I9M3
I. General Wholesale Prices ........... Relative to I890-99 -2.I0 I317 + 2.60 96.3 5.20 3.65 2. Gross Receipts of Railroads ........ $Iooooooo +3.6I 6I.i +I2.4 II2. 6.47 8.25
3. Net Earnings of Railroads ......... $IOOOOOOO + .6i6 25.2 + 3.23 39.7 2.o8 6.46 4. Exports of Mdse.. $iOooo0ooo + .846 74.7 + 7.2I III. 7.8I I3.5
I. General Wholesale Prices ........... ........ .48 2.9 .... .... I2.7
i6. Indices of Crop Yield per Acre.+ .432 95-3 .... | *** 7.8
* These series were used in the article by the writer entitled " Construction of a Business Barometer Based upon Annual Data," American Economic Review (December, i9i6) 6, pp. 739-769. The original sources of data are: (i) Aldrich Report; Bureau of Labor Statistics; (2, 3, W4) Poor's Manual of Railroads; (4, 5) Statis-
* These series were used in the article by the writer entitled " Construction of a Business Barometer Based upon Annual Data," American Economic Review (December, i9i6) 6, pp. 739-769. The original sources of data are: (i) Aldrich Report; Bureau of Labor Statistics; (2, 3, W4) Poor's Manual of Railroads; (4, 5) Statis-
tics for United States and other official reports; (6) Iron Age; (7, 8, 9) Statistical Abstract for United States; (Ic, II, I2, I3) Commercial and Financial Chronicle; (I 5) Dun (Average of Dun and Bradstreet since i902); (i6) Moore's Economic Cycles; (I 7) Statistics for United States.
t Standard deviation for period i874-96.
tics for United States and other official reports; (6) Iron Age; (7, 8, 9) Statistical Abstract for United States; (Ic, II, I2, I3) Commercial and Financial Chronicle; (I 5) Dun (Average of Dun and Bradstreet since i902); (i6) Moore's Economic Cycles; (I 7) Statistics for United States.
t Standard deviation for period i874-96.
few more items. The use of the geometric or arithmetic averages, on the other hand, would be apt to lead us into the false security of spuriously accurate decimal places.
few more items. The use of the geometric or arithmetic averages, on the other hand, would be apt to lead us into the false security of spuriously accurate decimal places.
3. GENERAL CONSIDERATIONS RELATING TO SECULAR TREND, I903-i8
Study of seventeen series of annual items for the period i879-I9I3 led to the conclusion that the secular trend of all the series was quite different during the subperiod i879-96 from that during the subperiod i897-I9I3. Straight lines fit the data very closely for the two subperiods but not for the whole period i879- I9I3. In Chart B illustration is given of the statement just made for the seventeen series. It will be observed from that chart that a single straight line does not fit the original series of per capita pig-iron production for the whole period, i86o-i9I5. If, however, we had broken the period into two subperiods, the first end- ing in the go's and the other beginning there, two straight lines would have shown the secular trend of each subperiod accurately. It can be seen by inspec- tion that the trend since the go's has been more strongly upward than it was before that decade. This statement applies without qualification to every one of the seventeen series investigated. In nine cases a downward slope for i879-96 became an upward slope
3. GENERAL CONSIDERATIONS RELATING TO SECULAR TREND, I903-i8
Study of seventeen series of annual items for the period i879-I9I3 led to the conclusion that the secular trend of all the series was quite different during the subperiod i879-96 from that during the subperiod i897-I9I3. Straight lines fit the data very closely for the two subperiods but not for the whole period i879- I9I3. In Chart B illustration is given of the statement just made for the seventeen series. It will be observed from that chart that a single straight line does not fit the original series of per capita pig-iron production for the whole period, i86o-i9I5. If, however, we had broken the period into two subperiods, the first end- ing in the go's and the other beginning there, two straight lines would have shown the secular trend of each subperiod accurately. It can be seen by inspec- tion that the trend since the go's has been more strongly upward than it was before that decade. This statement applies without qualification to every one of the seventeen series investigated. In nine cases a downward slope for i879-96 became an upward slope
in the succeeding period; in eight cases an upward slope for the earlier period became more pronounced in the later periods.' Table H gives the constants for the lines of secular trend, i879-96 and i897-I9W3, for the seventeen series studied.
In view of the evidence secured from the study of secular trend for the period i879-19I3, it appeared that the secular trend for any period beginning with the middle go's or later might safely and accurately be represented by a straight line. The available evidence indicates that had any other initial year been used, not earlier than the middle go's, the secular trend se- cured would not have been changed materially. Con- sequently, a straight line was used to represent the secular trend for the monthly data beginning, in most cases, with the year i903.
In the majority of the fifteen monthly series, consti- tuting the data for the present application of method, the terminal year of the period used for determining secular trend was i9i6. This year was. taken not only because it was the year previous to the entry of the United States into the Great War, but also because the years of prosperity, I9I5 and i9i6, were preceded by years of depression. In other words, in measuring
in the succeeding period; in eight cases an upward slope for the earlier period became more pronounced in the later periods.' Table H gives the constants for the lines of secular trend, i879-96 and i897-I9W3, for the seventeen series studied.
In view of the evidence secured from the study of secular trend for the period i879-19I3, it appeared that the secular trend for any period beginning with the middle go's or later might safely and accurately be represented by a straight line. The available evidence indicates that had any other initial year been used, not earlier than the middle go's, the secular trend se- cured would not have been changed materially. Con- sequently, a straight line was used to represent the secular trend for the monthly data beginning, in most cases, with the year i903.
In the majority of the fifteen monthly series, consti- tuting the data for the present application of method, the terminal year of the period used for determining secular trend was i9i6. This year was. taken not only because it was the year previous to the entry of the United States into the Great War, but also because the years of prosperity, I9I5 and i9i6, were preceded by years of depression. In other words, in measuring
I The upward slope of business failures during i879-96 and the downward slope during i897-I9I3 should be reversed, since the series is used as an index of business activity.
I The upward slope of business failures during i879-96 and the downward slope during i897-I9I3 should be reversed, since the series is used as an index of business activity.
This content downloaded from 194.29.185.26 on Thu, 15 May 2014 21:25:44 PMAll use subject to JSTOR Terms and Conditions
INDICES OF BUSINESS CONDITIONS 39 INDICES OF BUSINESS CONDITIONS 39
TABLE G -EQUATIONS OF LINES OF SECULAR TREND FOR VARIOUS SERIES
SERIES EQUATIONS * UNTTS
i. New York Clearings ........................................... = + .30x + 6.o6 Billion dollars 2. Production of Pig Iron ......................................... y = + 95x + I46i Thousand gross tons 3. Outside Clearings ............................................. y = + .29x + 3.46 Billion dollars 4. Bradstreet's Prices ............................................ y = + .IIx + 7.94 One dollar 5. Imports......................................................y = + .70x + 8.05 Ten million dollars 6. Building Permits . .............................................. Y + .99x + 35.2 Million dollars 7. Railroad Gross Earnings ....................................... y = + 2.95X + 40.8 Million dollars 8. Shares Traded .............................................. y - 52x + i8.22 Million shares 9. Unfilled Orders U. S. S. C....................................... y + .22x + 3.86 t Million tons
IO. Tonnage Entered ................ .............................. y + 70x + I92I t Thousand net tons ii. Business Failures (Bradstreet) ......... ......................... Y + 54X + 727 One failure I2. Rate on IO Railroad Bonds ..................................... y + .o0x + 3.85 One per cent I3. Rate on 4-6 Months Paper ..................................... y .iox + 5.98 ? One per cent I4. Rate on 60-90 Day Paper ...................................... y .09x + 5.26 One per cent
I5. Rate on Call Loans ............................................ Not found One per cent
* Abscissas (x) measured in years from i903. ? The equation of the line fitted for the period I8go-i9i,5 is y =
t Abscissas (x) measured in years from i908. -.05x + 6.29, where the abscissas (x) are measured in years from T Abscissas (x) measured in years from i900. i890.
TABLE G -EQUATIONS OF LINES OF SECULAR TREND FOR VARIOUS SERIES
SERIES EQUATIONS * UNTTS
i. New York Clearings ........................................... = + .30x + 6.o6 Billion dollars 2. Production of Pig Iron ......................................... y = + 95x + I46i Thousand gross tons 3. Outside Clearings ............................................. y = + .29x + 3.46 Billion dollars 4. Bradstreet's Prices ............................................ y = + .IIx + 7.94 One dollar 5. Imports......................................................y = + .70x + 8.05 Ten million dollars 6. Building Permits . .............................................. Y + .99x + 35.2 Million dollars 7. Railroad Gross Earnings ....................................... y = + 2.95X + 40.8 Million dollars 8. Shares Traded .............................................. y - 52x + i8.22 Million shares 9. Unfilled Orders U. S. S. C....................................... y + .22x + 3.86 t Million tons
IO. Tonnage Entered ................ .............................. y + 70x + I92I t Thousand net tons ii. Business Failures (Bradstreet) ......... ......................... Y + 54X + 727 One failure I2. Rate on IO Railroad Bonds ..................................... y + .o0x + 3.85 One per cent I3. Rate on 4-6 Months Paper ..................................... y .iox + 5.98 ? One per cent I4. Rate on 60-90 Day Paper ...................................... y .09x + 5.26 One per cent
I5. Rate on Call Loans ............................................ Not found One per cent
* Abscissas (x) measured in years from i903. ? The equation of the line fitted for the period I8go-i9i,5 is y =
t Abscissas (x) measured in years from i908. -.05x + 6.29, where the abscissas (x) are measured in years from T Abscissas (x) measured in years from i900. i890.
trend it was thought undesirable to include the years of depression without also including offsetting years of activity. The secular trend of wholesale prices was determined for the period ending in I914, because of the unprecedented disturbance of prices after that year.
In estimating the secular trend for the period since i9i6 (or, in the case of prices, since 1914) we have simply extended the straight lines determined from the data of the preceding period, I903-16. This has been done because the upheaval during the years I917, i9i8
is not thought to be in the nature of a change of trend. It is possible, indeed, that the secular trend in the decade beginning with i919 may be more strongly upward than it was in the period I903-i6, but more evidence is needed before such a conclusion can satis- factorily be maintained. Indeed evidence may be of- fered indicating that the trend of the immediate future will be less strongly upward than it has been. In any case, it is improbable that a change of trend, if a change has occurred, would be as strongly upward as would be the trend obtained by using straight lines determined from data including items for I9I7 and i9i8.
trend it was thought undesirable to include the years of depression without also including offsetting years of activity. The secular trend of wholesale prices was determined for the period ending in I914, because of the unprecedented disturbance of prices after that year.
In estimating the secular trend for the period since i9i6 (or, in the case of prices, since 1914) we have simply extended the straight lines determined from the data of the preceding period, I903-16. This has been done because the upheaval during the years I917, i9i8
is not thought to be in the nature of a change of trend. It is possible, indeed, that the secular trend in the decade beginning with i919 may be more strongly upward than it was in the period I903-i6, but more evidence is needed before such a conclusion can satis- factorily be maintained. Indeed evidence may be of- fered indicating that the trend of the immediate future will be less strongly upward than it has been. In any case, it is improbable that a change of trend, if a change has occurred, would be as strongly upward as would be the trend obtained by using straight lines determined from data including items for I9I7 and i9i8.
4. SUMMARY OF RESULTS FOR EACH SERIES.
The source and nature of the series presented graphi- cally in Charts i to I 5 are given in the following sum- mary. The numbers of the series (in parenthesis) are the same as those of the charts and frequency tables constructed from such series. Tables of the original series, the moving averages, the month-to-month link relatives, and the percentage deviations of the original items from the corresponding ordinates of secular trend corrected for seasonal variation are given in connection with Charts i to I5. The percentage deviations, just named, are presented graphically in Charts IoI to II4,
the units used being the respective standard deviations; the digits, in the units and tens places of the chart numbers, correspond to the number of the series upon
4. SUMMARY OF RESULTS FOR EACH SERIES.
The source and nature of the series presented graphi- cally in Charts i to I 5 are given in the following sum- mary. The numbers of the series (in parenthesis) are the same as those of the charts and frequency tables constructed from such series. Tables of the original series, the moving averages, the month-to-month link relatives, and the percentage deviations of the original items from the corresponding ordinates of secular trend corrected for seasonal variation are given in connection with Charts i to I5. The percentage deviations, just named, are presented graphically in Charts IoI to II4,
the units used being the respective standard deviations; the digits, in the units and tens places of the chart numbers, correspond to the number of the series upon
which they are based. (For instance, Chart i, Table i,
Chart ioi all pertain to Series i.) The equations of the lines of secular trend and the standard deviations are given in Tables F and G respectively. The indices of seasonal variation are given in the frequency tables of link relatives; graphs of the indices are presented as inserts of Charts i to I5.
which they are based. (For instance, Chart i, Table i,
Chart ioi all pertain to Series i.) The equations of the lines of secular trend and the standard deviations are given in Tables F and G respectively. The indices of seasonal variation are given in the frequency tables of link relatives; graphs of the indices are presented as inserts of Charts i to I5.
(I) MONTHLY BANK CLEARINGS OF NEW YORK CITY
SOURCE AND NATURE OF DATA. In the first or second issue each month of the Commercial and Financial Chronicle, monthly items of bank clearings for New York City appear under the caption " Bank Clearings at Leading Cities." This table gives a comparison of the month just completed with the corresponding month for three or more years preceding. Our figures have been secured from this source; but, when revised figures appeared, the current item in each issue has been re- placed by the item published a year later.
The tables presented by the Chronicle give both the absolute figures and the relative per cent of increase or decrease of the figures for the months of the current year over those of the preceding year; but only the absolute figures have been utilized. It is possible that the inauguration of the clearing system of the Federal Reserve Board has made New York clearings of little value as an index of business conditions. This system appears to have affected New York clearings much more than outside clearings. The effect on the latter is not evident. New York clearings reached a maximum in the autumn of i9i6; outside clearings are now, in the autumn of i9i8, establishing new high records.
SECULAR TREND. The linear trend, I903-16 (Chart
i), has a uniform annual increment of $295,000,000.
The average of the monthly items is $7,980,000,000.
The fit of the line to the data is close throughout, except during the war period, I9I4-I8.
(I) MONTHLY BANK CLEARINGS OF NEW YORK CITY
SOURCE AND NATURE OF DATA. In the first or second issue each month of the Commercial and Financial Chronicle, monthly items of bank clearings for New York City appear under the caption " Bank Clearings at Leading Cities." This table gives a comparison of the month just completed with the corresponding month for three or more years preceding. Our figures have been secured from this source; but, when revised figures appeared, the current item in each issue has been re- placed by the item published a year later.
The tables presented by the Chronicle give both the absolute figures and the relative per cent of increase or decrease of the figures for the months of the current year over those of the preceding year; but only the absolute figures have been utilized. It is possible that the inauguration of the clearing system of the Federal Reserve Board has made New York clearings of little value as an index of business conditions. This system appears to have affected New York clearings much more than outside clearings. The effect on the latter is not evident. New York clearings reached a maximum in the autumn of i9i6; outside clearings are now, in the autumn of i9i8, establishing new high records.
SECULAR TREND. The linear trend, I903-16 (Chart
i), has a uniform annual increment of $295,000,000.
The average of the monthly items is $7,980,000,000.
The fit of the line to the data is close throughout, except during the war period, I9I4-I8.
This content downloaded from 194.29.185.26 on Thu, 15 May 2014 21:25:44 PMAll use subject to JSTOR Terms and Conditions
