| 王守坤,汪柳.全国统一大市场发展进程的再测度[J].数量经济技术经济研究,2026,(10):242-268 | | 全国统一大市场发展进程的再测度 | | Remeasuring the Development of China’s Unified National Market | | | | DOI： | | 中文关键词: 全国统一大市场 市场分割 相对价格法 误区纠正 | | 英文关键词: Unified National Market Market Segmentation Relative Price Approach Misconception Correction | | 基金项目: | | | 中文摘要: | | 准确测度全国统一大市场发展进程，是评估其优化资源配置效率、推动经济高质量发展作用的重要前提。基于“一价定律”和“冰山成本”理论，通过商品相对价格差异衡量统一大市场发展水平，已成为当前广泛使用的测度方法。然而，在传统通用测算步骤中，存在将各类商品价格指数差异的绝对值直接去均值化即可消除商品自身属性的误解，以及误用跨商品大类的方差指标度量市场分割的局限。针对这些问题，本文认为两地区间商品相对价格差异其实已经包含了市场分割程度信息，并提出了基于广义自回归条件异方差（GARCH）模型剔除非政策性市场分割因素引起的相对价格差异的思路，以对传统相对价格法进行改进。基于全国251个城市、八大类商品相对价格指数的测度结果显示，传统方法测度的全国统一大市场建设进程变化率的波动幅度相对明显。进一步，放弃算术平均方法进行城市层面汇总，即采用加权方法分析发现城市间经济联系有必要在市场一体化进程测度步骤中加以体现。当前，对于商品相对价格法在测度过程中的应用误区尚缺乏广泛关注，本文为更加准确地衡量中国统一大市场发展进程提供了新的改进思路与客观审视。 | | 英文摘要: | | Accurately measuring the development of China’s unified national market is a prerequisite for evaluating its role in improving resource allocation efficiency and promoting high-quality economic development. Grounded in the Law of One Price and the theory of iceberg trade costs, the relative-price approach, which gauges market integration through cross-regional differences in commodity prices, has become a widely used method for assessing the development of the unified national market. However, the conventional procedure in the existing literature, which relies on differences in relative price indices across commodity categories, involves several methodological misconceptions.
First, the common practice of subtracting the national mean of the absolute relative price difference for a given commodity category from the corresponding bilateral absolute difference cannot, as often claimed, eliminate the intrinsic characteristics of that commodity category. Consequently, this transformation does not render relative price differences across commodity categories dimensionless or directly comparable, making subsequent variance calculations inappropriate. Second, the conventional practice of computing the absolute differences in relative price indices across multiple commodity categories for a given pair of regions and then taking the variance of this cross-category series lacks a well-defined economic interpretation. More specifically, and closely related to the first issue, because existing approaches do not adequately remove commodity-specific dimensional and distributional heterogeneity, price indices for different commodity categories are inherently drawn from different distributions. It is therefore methodologically questionable to calculate the variance across the absolute relative price differences of heterogeneous commodity categories.
Building on the identification of these problems, this paper re-examines the standard relative-price procedure for measuring the development of China’s unified national market and proposes a new approach based on a GARCH framework to filter out relative price differentials attributable to non-policy sources of market segmentation. This constitutes the central contribution of the paper. The GARCH framework is particularly well suited to the research question for two reasons. On the one hand, its underlying logic is to decompose fluctuations in a variable into an “expected” component and an “unexpected” component. The expected component reflects the persistent or long-memory part of the process, whereas the unexpected component captures short-run shocks. On the other hand, objective sources of market frictions, such as geographical distance, transportation costs, and other forms of iceberg trade costs, tend to be relatively stable over time. Market participants can therefore form persistent expectations regarding the relative price differentials generated by these factors, allowing such differentials to be interpreted as the expected component of relative price movements.
The applicability of the GARCH framework further requires the unexpected component to exhibit volatility clustering, conditional heteroskedasticity, and serial dependence. We argue that the component of relative price differentials generated by discretionary, policy-induced market segmentation satisfies these conditions. Local governments may strategically adjust the intensity and scope of market protection in response to changing economic conditions and regulatory pressure from higher-level authorities. Such adjustments are often embedded in interjurisdictional policy competition, strategic interaction, and imitation associated with the attraction of investment. As a result, policy-induced relative price differentials are likely to display pronounced volatility clustering, together with conditional heteroskedasticity and serial correlation.
Using relative price indices for eight major commodity categories across 251 Chinese cities, the results indicate that the conventional method yields relatively pronounced fluctuations in the rate of change of China’s unified national market development. We further depart from the arithmetic-mean aggregation commonly used at the city level and introduce a weighted aggregation scheme. The results indicate that intercity economic linkages should be explicitly incorporated into the measurement of market integration. By progressively varying the relative shares of economic and geographical components within a nested weighting structure, we are able to assess and validate the sensitivity of the measurement results to alternative weighting schemes. This provides a more comprehensive perspective for measuring the development of China’s unified national market.
At present, the methodological limitations of the relative-price approach have received insufficient attention in the literature. By identifying these application problems and proposing a GARCH-based correction, this paper offers a new methodological perspective for more accurately measuring and critically reassessing the development of China’s unified national market. | | 查看全文 相关附件： 下载数据代码附录 |
| | |
