投入产出表又称部门联系平衡表，最基本形式是一张将国民经济各生产部门排列在一起的矩阵表,反应一定时期内各部门间相互联系和平衡比例关系 ADDIN NE.Ref.{D036F048-A3D9-4560-98DC-C15046C389BE}^[1]，能够清晰地描绘产业经济系统结构，广泛应用于市场经济分析。自1936年Leontief提出并首次编制美国投入产出表以来，已有一百多个国家参与了编制。目前，我国已经能够在国家和省级层面制度性地开展投入产出表的编制工作，有规律地发布比较完善的投入产出表，为经济分析工作提供重要的数据基础，使得投入产出应用分析工作能够广泛开展。然而，在省级以下城市和地区层面投入产出数据还相当匮乏，一方面表现为官方组织基于调查法编制的城市投入产出表非常之少并且一般不对外公开发布，另一方面表现为其他组织、科研机构和个人也鲜有基于非调查法编制城市投入产出表的尝试。这导致目前的城市投入产出数据远远满足不了科研和实际工作的需求，长期以来构成限制我国开展投入产出应用分析工作的瓶颈。因此，研究、开发和编制城市投入产出表十分必要。

通过调查法编制城市投入产出表要求较大的人力物力和财力的投入，而非调查法不需要花费大量的人力物力，又能在保证一定程度精度的情况下，通过整合城市经济系统数据，为研究城市经济发展问题提供丰富的数据。因此本文的目标是尝试以武汉市为例，采用非调查法编制城市投入产出表并加以应用。

本文详细梳理了常用的非调查方法，包括区位商LQ及相关技术、RAS法、优化法及FES法等，作为本文的方法论核心。基于经济原理、所能获取的数据基础以及以往学者对各种非调查法的验证和评价，本文选用了区位商LQ法和RAS法相结合的方式编制武汉市投入产出表，期待本项工作能够为其他城市投入产出表的编制提供一定的借鉴意义。

本研究采用的编制方法由按比例估算和分解、总量控制、结构调整、分量平衡和系数分解等五种主要方法组成。首先，需要对投入产出表中各合计项进行整理或估算，这主要利用武汉市的总产值、增加值、消费、投资、进出口等总量数据，结合湖北投入产出表中的相应项来推算或估算获得。其次，结合已有数据基础以及这些估算得来的总量数据，参照一定的比例进行分行业分解，以获得各行业的总量数据。再次，依据湖北投入产出表的结构，应用区域投入产出表建造中常用的LQ区位商方法对中间使用系数矩阵和最终使用系数矩阵分别进行调整修正。最后，应用RAS法进行平衡处理，得到完整的武汉市投入产出表。

本文将利用编制出的投入产出表，在概述武汉市经济发展水平和产业结构的基础上，分析产业间的联动关系。一方面，通过对总产出、增加值、最终使用等总量数据和分行业总量数据的分析把握武汉市经济总量和产业经济结构，通过区位商分析评估武汉市各个产业部门的发展水平。另一方面，利用影响力系数和感应度系数分析产业间的拉动和驱动作用；利用Theil指数，测算武汉市的经济结构合理程度；利用生产诱发系数，分析各个产业最主要的最终需求驱动因素，以此来多角度地对武汉市产业间联动展开分析。最后针对性地提出了发展建议。

The input-output table, also known as the departmental balance table, is a balance table that reflects the interrelationships and balance ratios between various departments within a period of time. It can clearly describe the industry structure clearly and be used widely in economic analysis. Since Leontief proposed and compiled the US input-output table for the first time in 1936, more than 100 countries have participated in the compilation. At present, China has been able to compile and publish input-output tables at the national and provincial levels regularly. However, city-level input-output tables are still quite scarce in China. On the one hand, the city-level input-output tables compiled by official organizations are very few and not released publicly. On the other hand, there are few organizations or individuals have tried to compile the tables based on non-survey methods. The data are far from meeting the needs of scientific research, which has restricted related development. Therefore, it is necessary to research, develop and compile city-level input-output tables.

The compilation of the city input-output tables by survey methods requires a large amount of input, which are very scarce. The non-survey methods not only don’t need quite a lot of resources, but also can provide a certain degree of accuracy. Input output tables can provide a basis for quantitative analysis for relevant policy analysis. Therefore, the goal of this article is to try to use non-survey methods to compile and apply the city-level input-output tables using Wuhan as an example.

This article combs common non-survey methods in detail, including location quotient LQ and related technologies, RAS, optimization methods and FES, which are the core principles of this article. The paper selects the combination of location quotient LQ and RAS to compile the input-output table of Wuhan, based on economic principles, the available data, and verification and evaluation of various non-survey methods by previous scholars. It is expected that this work will be able to provide a reference for other cities.

We use five main methods: such as proportional estimation and decomposition, total control, structural adjustment, component balance and coefficient decomposition. First of all, it is necessary to sort or estimate the total items in the input-output table. We can use total output value, added value, consumption, investment, import and export and other total data in Wuhan, combined with the input-output table of Hubei. Secondly, decomposing the total data into each industry by a certain ratio. Third, according to the structure of the input-output table in Hubei, using LQ to adjust the intermediate use coefficient matrix and the final use coefficient matrix respectively. Finally, applying the RAS to balance data in order to obtain a complete input-output table of Wuhan.

This essay will analyze the industrial linkage on the base of summary of Wuhan's economic level and industrial structure. On the one hand, through the analysis of total data and sub-industry total data, such as output, added-value, final usage, grasping the economic aggregate and industrial economic structure in Wuhan. And we evaluate the development level of various industrial sectors by location quotient analysis. On the other hand, using the influence coefficient and the inductance coefficient to analyze the driving and pulling effects between industries; using the Theil index to measure the rationality of Wuhan’s economic structure; using the production induction coefficient to analyze the most important final demand drivers of each industry. In this way, we analyze the linkages between industries in Wuhan from multiple angles. Finally, some suggestions were made.