中國GDP核算誤差的特征事實與發(fā)生機制——兼論計量模型方法對于GDP數(shù)據(jù)質(zhì)量的評估功效

摘要： 基于GDP數(shù)據(jù)的歷史修訂與地區(qū)差異情況,依照反事實分析的推理邏輯,考察GDP核算誤差的特征事實,探討其中覆蓋誤差、操作誤差以及估算誤差的發(fā)生機制與影響模式。結(jié)果表明,在各年份現(xiàn)價GDP數(shù)據(jù)中,由覆蓋誤差和估算誤差造成的綜合效應(yīng)是以系統(tǒng)性的低估為主,其中服務(wù)業(yè)部門是核算誤差的主要來源所在;地區(qū)層級GDP核算過程中覆蓋誤差、估算誤差與操作誤差的綜合效果以持續(xù)性的數(shù)據(jù)高估為主,其中的操作誤差效應(yīng)也會存在于國家層級的GDP核算中。由于缺乏對于GDP核算誤差發(fā)生機制的必要考慮,計量模型方法在評估GDP數(shù)據(jù)質(zhì)量中的功效非常有限。
The paper aims at probing into the generation mechanism of Chinese GDP accounting error,with a counterfactual analysis on official data revision and regional difference of GDP statistics.It can be found that the accounting error derived from imperfect survey coverage and improper estimation had a synthesized effect of systematic underestimation on GDP level,and at the same time,there had been a persistent overestimation of the national-level statistics caused by operational disturbance. Without considering relevant generation mechanism of Chinese GDP accounting,those evaluations on the accuracy of Chinese GDP statistics based on specific econometric models can hardly realize remarkable effects.