Freight rates and shipping costs in the English coastal coal trade: a reply
本文回应了Armstrong对作者此前航运生产率研究的批评，指出其方法高估了运费并低估了生产率变化，同时提供了基于Harley和议会文件的新运费序列。
In my article on shipping productivity in the coal trade between the north east and London, shipping costs were calculated by taking the London pool price, and subtracting the price of coal in the north east and the coal tax. These figures were then analysed as if they were equivalent to freight rates.' Armstrong is correct to say that this methodology led me to overestimate 'freight rates' at the end of the period substantially.2 The most serious consequence of this was an underestimation of productivity change in shipping (as measured by the inverse of the freight rate index) from around i875 to i890, the period during which the adoption of large steam colliers had an obvious impact on the trade. Armstrong's figures, based on published freight rates, are superior to my estimates, but he has missed a freight rate series by Harley that uses sources similar to his, but extended from i838 to I9I3.3 When combined with figures published in Parliamentary Papers, a direct series for freight rates can be extended back to i8i8. Figure i presents this series along with my calculated series. In most years my figures are higher than the freight rate figures, and they are more volatile.4 Table i compares estimates of annual growth rates for the two series. Overall, the freight rate series falls 2.5 times faster than mine, but most of this discrepancy can be attributed to the years from about i875 to I890.5 For the period covered by Armstrong, the rate of decline in the two series is nearly identical. My series differs from the freight rate series for two reasons: first, my calculated series almost certainly contains measurement errors; second, my series measures more than just bare freight costs. The most troublesome aspect of putting together the various series that comprised coal prices in