Click the image to purchase this issue
Abstract:Long-term historical statistics are an important foundation for the West to construct modern economic growth theories and ensure the discourse power of the Western modernization path. Since the mid-20th century, in the close interaction between policy demands in Western countries and academic supply, various Western countries have undergone a large-scale rise and transformation of long-term historical statistics, especially in the past twenty years, where the pace of building various global long-term historical databases has accelerated. Studying the background, evolution stages, research content, development trends, and academic impacts of the rise of long-term historical statistics in the West is of significant reference for constructing a long-term historical statistics database with Chinese characteristics today, supporting the continuous improvement of the theoretical system of Chinese-style modernization.
Keywords:Long-term Historical Statistics, Modern Economic Growth, Historical Database
Authors:XU Yi, Professor at the College of History, Culture, and Tourism, Guangxi Normal University; ZHUANG Zicheng, PhD student at the College of History, Culture, and Tourism, Guangxi Normal University. (Guilin 541001)
Source:“Chinese Social Science Evaluation”, 2024, Issue 1, P66—P76
Editor:LIANG Hua
In the Western modern social science system, the theory of modern economic growth is one of the foundational theories for constructing the Western modernization model. In the 1930s and 1940s, American scholars such as Simon S. Kuznets relied entirely on tracking, depicting, and studying the modern economic growth of developed Western countries to construct the theory of modern economic growth, narrating the economic growth stories of developed Western countries through quantification. Therefore, collecting various data and quantifiable materials as much as possible, compiling them into series to depict the characterized facts of modern economic growth, has become an important task for the Western economic and economic history communities. In fact, with the official publication of various economic statistics by Western countries in the 19th century, various academic organizations attempted to compile the official statistical data of various countries. However, the truly significant long-term historical statistics published publicly, which involves collecting and organizing various indicators reflecting economic growth and social transformation over a long period (over 100 years) and constructing comparable series, began in the mid-20th century. Since then, Western countries have generally experienced a large-scale rise and transformation of long-term historical statistics, which has played an important supporting role in the Western academic community’s construction, development, and improvement of modern economic growth theories, maintaining the discourse power of the Western modernization path. Studying the background, evolution stages, research content, development trends, and academic impacts of the rise of long-term historical statistics in the West is of significant reference for constructing a long-term historical statistics database with Chinese characteristics today, supporting the continuous improvement of the theoretical system of Chinese-style modernization.
During World War II, to better mobilize their economic resources, Western countries began research on long-term economic fluctuations and national income accounting. By the mid-1940s, the capitalist economy in the West entered a critical development period. In the United States, economists were concerned that the demobilization of over 20 million military personnel would trigger a new wave of Great Depression in the U.S., and they were continuously studying and exploring public policies to ensure post-war economic growth and stability. In Europe, economists were concerned about reviving the war-torn economy, actively exploring public policies to restore Europe’s traditional advantages and competitiveness. Consequently, the Western academic community saw a surge in long-term growth theory and empirical research. To support the Western academic community’s research on long-term economic growth, since the mid-20th century, Western countries such as the United States, Italy, the United Kingdom, Canada, New Zealand, Norway, and Sweden have successively initiated the collection, organization, and compilation of their long-term historical statistical data. Initially, the long-term historical statistics in Western countries were derived from the annual data compilations published by their respective statistical departments. Among these countries, the United States was the first to begin compiling long-term historical statistics. Influenced by the U.S., the U.K. compiled the most comprehensive long-term historical statistics at that time; while Italy became the first country on the European continent to compile long-term historical statistics. We focus on elaborating the processes of these three countries in compiling their long-term historical statistics to reflect the characteristics of the rise of long-term historical statistics compilation in the West.
(1) Compilation Entities
Since the 1940s, the compilation of long-term historical statistics in Western countries has been undertaken by their national statistical agencies. In the U.S., which was the earliest to compile long-term historical statistics, this was initiated by the Statistical Reporting Division of the U.S. Census Bureau. Since 1878, the U.S. federal government has published the “Statistical Abstract” annually. In 1943, this work began to be managed by the Statistical Reporting Division of the Census Bureau, which proposed to supplement and extend the statistical scope and time span of the “Statistical Abstract” to meet the growing demand for long-term historical data from economists and sociologists at that time, planning to publish a compilation of U.S. long-term historical statistics. Meanwhile, the Social Science Research Council (SSRC), which brought together experts from various fields of social science, was also conducting important projects related to long-term growth, including analyzing the development trends of indicators such as U.S. production, income, and consumption, thereby assessing the economic potential of the U.S. and exploring its economic growth drivers. This project required long-term historical statistical data as support. Due to the same demand and respective advantages, the Census Bureau and the SSRC collaborated in the late 1940s, drawing on the methods used by the National Bureau of Economic Research (NBER) to construct time series data, and in 1949 published the first compilation of long-term historical statistics in the U.S. – “Historical Statistics of the United States (1789-1945).”
Similarly, in Italy, the government has always placed great importance on statistical work. Since 1861, Italy has required different departments to carry out statistical work, continuously accumulating statistical data. By 1926, the Italian Central Statistical Institute was officially established. After long-term investigation, research, and study, they cultivated a group of professionals engaged in statistical work. In 1955, marking the 30th anniversary of the institute, they planned to compile a long-term historical statistics compilation, partly to review the arduous path of Italian official statistics over the past century; and also to showcase Italy’s economic growth and social transformation from 1861 to the 1950s to the academic community and the public through the compilation of long-term historical data. Based on this, the Italian Central Statistical Institute organized personnel to compile and publish Italy’s first long-term historical statistics compilation – “Statistical Abstract of Italy (1861-1955)” in 1958.
Unlike the U.S. and Italy, in the U.K., the compilation of long-term historical statistics was led by a research team headed by Cambridge University Professor B. R. Mitchell, commissioned by the Social Science Research Council of the Department for Education and Science. Influenced by the U.S., both the political and academic circles in the U.K. recognized the importance of compiling long-term historical statistics. Consequently, Mitchell led a team composed of statisticians from various government statistical departments, universities, and research institutions, working together to publish the first compilation of long-term historical statistics in the U.K. – “Statistical Abstract for the United Kingdom” in 1962. Initially, the government departments in these three countries decided to roughly recompile their long-term historical statistics every ten years. By the mid-1970s, the U.S. had compiled three editions, Italy had compiled three editions, and the U.K. had compiled two editions.
In summary, in the early stages of compiling long-term historical statistics, the work was primarily initiated and organized by government departments in each country, with only a small number of scholars participating.
(2) Data Sources
Since the compilation entities are national statistical agencies or teams composed of statistical agency personnel and scholars, the long-term historical statistics compilations in Western countries were primarily based on official annual statistical data, with only a very small part of the data derived from research by scholars. For example, the 1949 edition of “Historical Statistics of the United States (1789-1945)” primarily relied on five categories of U.S. official statistics: the first category consists of comprehensive statistical materials compiled annually by the U.S. federal government, such as the “Statistical Abstract”, which contains relatively comprehensive data. The second category includes annual work reports from U.S. federal government officials or departments, such as the “Annual Report of the Director of the Mint”, “Annual Report of the Commissioner of Immigration”, “Annual Report of the Secretary of Labor”, “Annual Report of the Commissioner of the General Land Office”, “Annual Report of the Secretary of the Treasury”, “Census Bureau Work Reports”, “Navy Department Reports”, “Annual Reports of the Comptroller of the Currency”, etc. These materials record the data related to the work under the jurisdiction of different departments, such as the number of immigrant population, the amount of vacant public land, tax revenue, fiscal expenditures, etc. The third category consists of census reports that record the development status of various industries, such as “Agricultural Census Reports”, “Electric Power Industry Census Reports”, “Mining Survey Reports”, etc. The fourth category includes specialized statistical materials formed through surveys and calculations by government departments, such as the Census Bureau’s “Population Census Reports”, “Birth Statistics Reports”, the U.S. Public Health Service’s “U.S. Population Dynamics Statistics”, the U.S. Bureau of Labor Statistics’ “Manufacturing Statistics Manual”, the Department of the Interior’s Bureau of Mines’ “Mineral Yearbook”, etc. The fifth category consists of various annual work reports submitted by enterprises to federal government departments, such as annual reports submitted by telegraph companies to the Interstate Commerce Commission.
The 1962 edition of “Statistical Abstract for the United Kingdom” is similar to “Historical Statistics of the United States (1789-1945)” in that it also compiled five categories of official statistics. The first category consists of comprehensive statistical handbooks published by the U.K. government, such as the “Annual Statistical Abstract”. The second category includes work reports from government officials or departments, such as the “Annual Report of the Secretary of Scotland”, “Annual Report of the Director of the Tax Office”, “Report of the Patent Commissioner”, annual reports from the Ministry of Mines, reports from the Special Committee on Smuggling, etc. The third category includes statistical data from industries, such as “Irish Agricultural Statistics”, the “Statistical Yearbook” of the British Iron and Steel Federation, etc. The fourth category consists of specialized statistical materials compiled by government departments, such as the “Population Census Table” and “Annual Trade Reports”. The fifth category includes industry reports from government meeting documents, similar to reports from the “Silk Trade Special Committee” and the “Hand Loom Weavers Special Committee”.
Compared to the U.S. and the U.K., Italy used fewer types of official materials to compile long-term historical statistics, which can be roughly divided into three categories. The first category includes data recorded by government departments in their work, such as records of fishery product trade by the Ministry of Agriculture and Forestry, tariff data recorded by the Ministry of Finance, and data on paper products recorded by the Ministry of Industry and Trade. The second category includes work reports and published bulletins from government officials or departments, such as the “Report of the Minister of Primary Education” and the “Official Bulletin of the Ministry of Education”. The third category consists of specialized statistical materials formed through surveys and calculations by government departments, such as the “Report on Italian Book Production” from the Ministry of Education, the “Statistical Data on Italian Silkworm Production” from the Ministry of Industry and Commerce, and the “Conversion Factor of the Lira from 1871 to 1952” published by the Central Statistical Office, etc.
In conclusion, the early compilations of long-term historical statistics in Western countries primarily utilized official statistical data from the 19th century onwards; while the statistical data prior to the 19th century were referenced from the quantitative research results of economic historians, but this portion of data occupied a very small space. For example, the national income series for the U.S. from 1863 to 1935 cited in “Historical Statistics of the United States (1789-1945)” came from the relevant research of American scholar Rufus S. Tucker, and the national income series for the U.K. from 1688 to 1955 cited in “Statistical Abstract for the United Kingdom” came from the research of British scholar Phyllis Deane.
(3) Statistical Content and Methods
In terms of statistical content, the statistical themes designed in the long-term historical statistics compilations of Western countries mainly focus on the performance, characteristics, and international competitiveness of long-term economic growth in their countries. The 1962 edition of “Statistical Abstract for the United Kingdom” contains the most comprehensive themes and covers the longest period, specifically including nearly 190 indicators across about 4000 series, covering themes such as population and life statistics, labor force, agriculture, coal, steel, tin, copper, lead, textile industry, transportation, construction, foreign trade, miscellaneous production, wages and living standards, national income and expenditure, public finance, banking and insurance, prices, etc. Although the 1949 edition of “Historical Statistics of the United States (1789-1945)” contains fewer themes than the U.K., some themes are more macro, covering over 80 indicators across about 3000 series from 1789 to 1945, including themes such as national wealth and income, population characteristics and migration, life cycle, health and nutrition, labor force, wages and working conditions, agriculture, land, forestry and fisheries, minerals and electricity, construction and housing, manufacturing, transportation, price indices, balance of payments and foreign trade, banking and finance, government, etc. In comparison, the 1958 edition of “Statistical Abstract of Italy (1861-1955)” includes fewer themes, but still reflects the long-term growth of major economic sectors in Italy from 1861 to 1955, involving 28 indicators across over 2000 series related to population, cultural and social activities, agriculture and forestry, industry, transportation and communication, foreign trade, credit and currency, prices and wages, public administration and national economic budget, etc.
Other Western countries, like the aforementioned three, published the first batch of long-term historical statistics indicators primarily based on data related to their economic and social development since the 18th and 19th centuries, with very few statistical data prior to the 19th century. Moreover, the indicators reflecting the performance, stages, and paths of modern economic growth were not comprehensive enough, and the indicators revealing the driving forces behind modern economic growth were even scarcer, while the indicators related to social transformation triggered by economic growth were relatively singular.
In terms of statistical methods, before the 1970s, the main method used by Western countries to compile long-term historical statistics was to directly obtain existing official statistical materials and summarize the data according to the designed indicators, at most using simple estimation methods to supplement the missing data. The most common practice was to use simple linear interpolation to supplement missing data for intermediate periods. For example, in the 1949 edition of “Historical Statistics of the United States (1789-1945)”, the editors estimated the population growth rate of the U.S. from 1900 to 1919 using data from the U.S. School Population Census, state population censuses, and other local data reflecting population changes, and combined it with known data on population gender, age, and race to estimate the missing data for intermediate years. Another common method was to calculate the proportion of missing data for a specific indicator based on the data from previous and subsequent years, and then infer the missing data for those years. For instance, in the 1962 edition of “Statistical Abstract for the United Kingdom”, when estimating income tax for the years 1843-1939, since life insurance premiums were approved for deduction before tax during the period from 1854 to 1874, but there were no related records of life insurance premiums, statisticians estimated the life insurance premiums for that period based on the proportion of life insurance premiums to total income in 1875, and then inferred the income tax after deducting life insurance premiums. In conclusion, the methods used to construct the first batch of published long-term historical statistics were relatively simple, lacking the use of more complex statistical methods and econometric approaches to construct long-term historical series.
After the publication of the first batch of long-term historical statistics compilations represented by the U.S., U.K., and Italy, there was a widespread interest in long-term historical statistical data across various sectors in the West. These compilations of long-term historical statistics not only became essential tools for economic historians and educators, but were also cited hundreds of times annually in academic, professional, and news publications. This series of chain reactions sparked great research interest in long-term historical statistical data within the Western academic community, leading to the emergence of many specialized fields focusing on constructing long-term historical statistical data, such as studies on wages, occupational structures, national accounts systems, and import-export trade in various Western countries. More importantly, the first batch of long-term historical statistics compilations strongly supported the research of modern economic growth theories in the West. During the compilation process of the first editions of long-term historical statistics in the U.S., U.K., and other countries, a number of influential theoretical achievements regarding modern economic growth were also published. As the U.S. and U.K. completed their first and second editions of long-term historical statistics, they also published a series of foundational theoretical results, such as British scholar Phyllis Deane’s “British Economic Growth: Trends and Structures (1688-1959)”, American scholar Douglass C. North’s series of results regarding U.S. economic growth in the 18th and 19th centuries, and the series of results by American scholar Simon S. Kuznets regarding modern economic growth in developed Western countries. These results provided empirical foundations, analytical frameworks, and theoretical perspectives for the eventual formation of modern economic growth theories in the West.
By the mid-1970s, Western long-term historical statistics underwent a transformation, shifting from compilations of second-hand materials by official agencies to research on various historical statistical data, that is, the collection, organization, comparison, accounting, and interpretation of data from multiple sources.This transformation has profound socio-economic and political backgrounds.Since the 1970s, the Western economy has entered a period of slow growth.Western countries urgently needed more comprehensive and systematic long-term historical statistical data to continuously improve modern economic growth theories in the West in order to seek new drivers for reviving economic growth in public policy and to maintain the Western advantage in modern economic growth theory.At the same time, the explosive growth of quantitative economic history research in the mid-20th century produced a large number of quantitative research results; coupled with the rapid development of econometrics and information technology, these all provided various opportunities and conditions for the transformation of Western long-term historical statistics during this period.
The first to undergo transformation were those countries that had already published long-term historical statistics compilations in the 1940s, such as the U.S., U.K., and Italy, which gradually shifted towards long-term historical statistical research led by scholarly teams starting in the mid-1970s. Subsequently, those Western countries that had previously not engaged in long-term historical statistical work also began to adopt the new model of scholarly team operations in the mid-1970s, publishing a series of long-term historical statistical research results. More importantly, since the 1990s, with the acceleration of globalization and the increasing economic strength of non-Western countries, Western countries needed a clearer understanding of the scope of economic growth experiences, catch-up processes, growth convergence and divergence, and their causes when formulating public policies. Therefore, under the funding of various governments or international organizations such as the OECD, Western academia formed multinational research teams and initiated global long-term historical statistical research projects to construct thematic and comprehensive global long-term historical statistical databases. The transformation of Western long-term historical statistics since the mid-1970s is specifically reflected in the following aspects.
(1) The Compilation Entities Shift from Statisticians to Scholars
Since the mid-1970s, scholars in disciplines such as economics and economic history in Western countries have increasingly participated in long-term historical statistical work, pushing the content of this work from compiling official statistical data to researching various data. Some scholars and statisticians continued to dedicate themselves to their country’s long-term historical statistical work. For example, in the U.S., although the 1949 edition of “Historical Statistics of the United States (1789-1945)” and the 1960 edition “Historical Statistics of the United States: Colonial Times to 1957” received wide recognition in the Western academic community, these two editions still had obvious limitations, as they did not cover research results in new fields related to the driving forces of economic growth, such as education, social security, and research and development. Therefore, the U.S. Census Bureau invited scholars to update the compilation of U.S. long-term historical statistical data. Nearly 130 experts and researchers from well-known universities and research institutions in the U.S. were invited to participate in this update effort, and after much work, the third edition of U.S. long-term historical statistics – “Historical Statistics of the United States: Colonial Times to 1970” was officially published in 1975. For the next 30 years, “Historical Statistics of the United States” was not updated. Entering the 21st century, the update of the long-term historical statistics compilation was put on the agenda again. However, unlike before, this time the compilation work was spontaneously organized by over 80 scholars from nearly 40 well-known universities and research institutions in the U.S., who contributed their research results on various data and published the five-volume “Historical Statistics of the United States (Millennium Edition)” in 2006. Thus, the subject of compiling U.S. long-term historical statistics completely shifted from government departments to scholars.
The transformation of long-term historical statistics in the U.K. mainly occurred in the 1980s. The long-term historical statistical work in the U.K. was originally commissioned by the Social Science Research Council of the Department for Education and Science to be completed by Professor Mitchell’s team at Cambridge University. Entering the 1980s, after more than a decade of development, many significant quantitative research results emerged in the U.K. economic history community, and Mitchell’s team invited over 20 scholars from well-known universities and research institutions in the U.K. to participate in this update effort. With the efforts of the scholars, the “Statistical Abstract for the United Kingdom” was published in 1988. In 2010, Mitchell’s team gathered nearly 60 well-known scholars in the U.K. again to update long-term historical statistical data and published the 2011 edition of “Statistical Abstract for the United Kingdom”.
Compared to the U.S. and U.K., the transformation of Italy’s long-term historical statistics compilation work was slower. By the 1980s, the Italian Central Statistical Institute began inviting scholars to participate in the long-term historical statistics update work, publishing “Statistical Abstract of Italy (1926-1985)” in 1986. By 2011, the Italian Central Statistical Institute invited over 20 scholars from well-known universities and research institutions in the country to work together with government statisticians to carry out long-term historical statistics updates, resulting in the publication of “Italy After 150 Years: Statistical Abstract (1861-2010)” in 2011.
In fact, this transformation occurred not only in the U.S., U.K., and Italy, but also in other Western countries’ long-term historical statistics, which experienced the aforementioned trend of transformation. To showcase the research results of long-term historical statistics transformation across European countries, Mitchell’s team collected and organized historical statistical data from over 20 European countries in 1975 and published “Historical Statistics of Europe (1750-1970)”. Since then, Western countries and international organizations have funded well-known scholars to form teams to engage in national, regional, and even global long-term historical statistical research.
(2) Data Sources Shift from Official Statistics to Various Types of Data
Before the mid-1970s, long-term historical statistics in Western countries primarily relied on official statistical data. After the mid-1970s, in order to further update long-term historical statistical data, countries began to collect data from various channels. With the involvement of scholars, the official statistics from the 18th-century industrial revolution onwards were no longer sufficient to meet the needs of scholars in various countries. On the one hand, scholars increasingly focused on indicators and data related to socio-economic development before the 18th-century industrial revolution; however, at that time, many countries had not yet established a system of socio-economic census, lacking systematic official statistical data, leading scholars to collect, organize, and construct relatively continuous series from various scattered materials. On the other hand, scholars also increasingly focused on the revision and supplementation of official statistical indicators and data after the 18th-century industrial revolution, or on new indicators, new issues, and new trends beyond official statistical indicators and data. Therefore, scholars either conducted their own research on these new aspects to construct new series or cited new research results from their peers. It can be said that the long-term historical statistics research teams in Western countries constructed long-term statistical series after the mid-1970s, citing a large number of quantitative research results from Western economic historians and attaching a wealth of original materials supporting these quantitative research results.
(3) Statistical Content Shift from Single National Themes to Global Comprehensive Themes
Long-term historical statistics in Western countries from the 1940s to the 1960s primarily focused on themes and indicators of long-term economic growth in their respective countries. After the mid-1970s, Western scholars began to pay attention not only to the performance, characteristics, and international competitiveness of their national economic growth indicators but also to the stages, paths, drivers of long-term economic growth, and the social transformations triggered by economic growth, as well as the economic growth and social transformation issues of non-Western countries. Meanwhile, the constructed series also expanded backward from the 18th-century industrial revolution.
The first changes were seen in the long-term historical statistics content published by Western countries. Specifically, the first type of change mainly involved adding themes related to economic growth paths, drivers, and social transformations. For example, compared to the 1949 edition of “Historical Statistics of the United States (1789-1945)”, the 1975 edition of “Historical Statistics of the United States: Colonial Times to 1970” added themes such as consumer spending and income, society, climate, communication, energy, services, enterprises, financial markets, productivity, and technological development. Similarly, compared to the 1958 edition of “Statistical Abstract of Italy (1861-1955)”, the 1976 edition of “Statistical Abstract of Italy (1861-1975)” added themes such as climate, health, education, justice, credit and insurance, and food consumption. The second type of change involved expanding and updating existing indicators within established themes. For instance, the 1975 edition of “Historical Statistics of the United States: Colonial Times to 1970” added the indicator and series of “Gross Domestic Product” under the original theme of “National Income and Wealth”. The 1971 edition of “Statistical Abstract for the United Kingdom 2” added the indicator series of “Average Weekly Wages of Manufacturing Workers” under the original theme of “Wages and Living Standards”. It is noteworthy that the long-term historical statistics results published after the mid-1970s have extended the constructed historical series back to the early modern period and even to the mid-Middle Ages. For example, the 2006 edition of “Historical Statistics of the United States (Millennium Edition)” traced back to the early 17th century; the 2011 edition of “Statistical Abstract for the United Kingdom” traced back to the 10th century.
After the 1980s, with the rapid development of emerging economies represented by China, India, and Brazil, the Western academic community increasingly focused on modern economic growth experiences from non-Western countries. Therefore, following the publication of “Historical Statistics of Europe (1750-1970)” by Mitchell’s team in 1975, they constructed long-term historical statistical series for other continents according to the thematic and indicator framework of Europe, publishing “World Historical Statistics: Asia and Africa” in 1982 and “World Historical Statistics: Americas and Oceania” in 1983. After the 1990s, Western scholars began to initiate projects to construct global long-term economic growth series. The earliest global project came from Dutch scholar Angus Maddison’s team, which constructed ultra-long series of population, GDP, and per capita GDP for over 160 countries from the year 1 AD to 2000. American scholar Peter H. Lindert initiated a project to construct global wage historical series, collaborating with British scholar Robert C. Allen, American-Chinese scholar Ma Debin, Dutch scholar Jan Luiten van Zanden, and others to construct historical series of real wages for over 90 countries from the 15th to the 20th century. German scholar Joerg Baten led a team dedicated to collecting, organizing, and comparing data on the heights of adults in various countries from the 19th to the 20th century, constructing historical series of adult male heights from 1810 to 1989 for 156 countries. American scholars Diego A. Comin and Bart Hobijn constructed historical series of the cross-national diffusion of over 100 technologies adopted by more than 150 countries since 1800, based on the global data from Mitchell’s team and combined with their own collection of data from other channels. All these global projects focused on specific themes and constructed specialized long-term series. Recently, Jan Luiten van Zanden initiated a comprehensive project for global long-term historical statistics, constructing series covering 210 countries across Europe, America, Asia, Oceania, and Africa since 1500, involving 78 categories of indicators related to themes such as agriculture, population, environment, finance, human capital, gender equality, institutions, labor relations, national accounts income, prices, wages, and minerals, totaling nearly 60,000 series.
(4) Research Methods Shift from Simple Data Compilation to Complex Series Construction
As mentioned above, before the 1970s, Western countries primarily compiled long-term historical statistics by directly obtaining existing official statistical materials and summarizing the data according to the designed indicators, at most using simple interpolation methods to supplement missing data. After the mid-1970s, with the development of econometrics and innovations in information science and technology, scholars increasingly introduced complex models, methods, and techniques into the construction of long-term historical statistical data for mining, organizing, comparing, accounting, interpreting, and networking the data.
First, Western scholars increasingly introduced various econometric models and methods to expand, construct, and improve long-term historical series. This method was mainly used to construct historical series for various themes such as population, employment, wages, prices, GDP, education, and welfare before the 18th century. Additionally, for the supplementation and improvement of historical series after the 18th century, scholars frequently employed econometric estimation methods. Furthermore, the Western academic community strengthened the interpretation and trend research of constructed series after the mid-1970s. The long-term historical statistics results published after this period, in addition to historical series, included textual materials at the beginning of each theme, detailing the meanings and sources of the data for each indicator and the methods used by scholars and statisticians in constructing long-term series, as well as precautions for using these data. Some countries even conducted preliminary studies on data changes and the trends exhibited by these changes. The 2006 edition of “Historical Statistics of the United States (Millennium Edition)”, the 2011 edition of “Statistical Abstract for the United Kingdom”, and Maddison’s “Millennial Statistics of the World Economy” all conducted in-depth trend research on the indicators and series of various themes.
More importantly, some scholars also utilized computer information processing technology to construct various long-term historical statistical databases. Starting in the 1980s, Mitchell’s team, with the technical support of Cambridge University Press, began to digitize the long-term historical statistics they published for each continent, constructing the first global long-term historical statistical database. Entering the 21st century, Maddison’s team constructed a global national accounts database, Lind’s team built a global price and wage historical database, van Zanden’s research team created the Clio-Infra database, and Comin and Hobijn’s technology cross-national diffusion historical database (CHAT), among others. These databases constructed by various countries not only contain all the data but also provide readers with downloadable Excel or CSV format data tables, allowing readers to edit data within these tables according to their needs. In the databases, readers can also conduct targeted searches for relevant data tables and accompanying documents related to the data. If data from the databases are used, readers can also download complete citations in RIS, text, or CSV format for annotation. These web-based databases have a data capacity and convenience that compilations cannot compare to.
In conclusion, since the transformation in the mid-1970s, teams of scholars in Western countries have combined historical data from various channels with complex econometric models and computer information processing technology, making the construction of global thematic or comprehensive historical databases a development trend in the field of long-term historical statistics. Not only have the long-term historical statistics results produced by scholarly teams in various countries become foundational data for economic history research in the Western academic community, but the research on long-term historical statistical data itself has also become an important component in Western economic history research. With the continuous promotion of scholarly research teams, in addition to periodic updates of national long-term historical statistics results, various important global long-term historical statistical databases will also be updated regularly. For example, Mitchell’s team completed updates of long-term historical statistical data for major continents in 2007; Maddison’s global national accounts database completed its most recent update in 2020; and van Zanden’s Clio-Infra database was expanded and updated in 2021. More importantly, these long-term historical statistical databases have significantly promoted the continuous expansion and innovation of modern economic growth theories in the West. First, in the last 30 years of the 20th century, the national long-term historical statistics results published by countries such as the U.K., U.S., Italy, and the Netherlands allowed the Western academic community to trace the origins of modern economic growth back to the early modern period and even to the late Middle Ages after the Black Death, providing a systematic data support for re-understanding the stages, manifestations, and driving forces of the origins of modern economic growth in the West. Secondly, since the 21st century, the Western academic community has utilized various global long-term historical statistical databases to construct a comprehensive indicator system to measure the changes in modern economic growth and welfare levels across countries globally since the 18th century; and through research on the modern economic growth paths of various major regions and countries globally, they are attempting to construct a modern economic growth theoretical framework that transcends the past experience of solely focusing on Western modern economic growth and holds global significance.
By reviewing the large-scale rise and transformation of long-term historical statistics in the West since the mid-20th century, revealing the close interaction between the policy demands of Western countries and the supply from academia, it can be seen that long-term historical statistics are an important foundation for Western countries to respond to economic crises, formulate public policies, construct, develop, and improve modern economic growth theories in theory, and maintain the discourse power of the Western modernization path. Especially since the 21st century, Western countries represented by the U.S. and Europe have accelerated their research on global long-term historical statistics and the construction of databases, insisting on seizing the “theoretical discourse power” in the competition for global economic development models, global modernization development paths, and global economic and political outlooks through the approach of “historical series construction – big data analysis – theoretical innovation”, with a clear trend of “database warfare” emerging.
We urgently need to base ourselves in China, learn from and adopt the beneficial experiences of Western long-term historical statistics research and database construction, and accelerate the construction of our own open long-term historical statistics database system. First, we should expedite systematic research on the origins, evolution, distribution, advantages, characteristics, research directions of Western long-term historical research and its databases, as well as their strategic impacts on the public policies of Western countries and international organizations, thereby establishing various channels and mechanisms to flexibly utilize Western database resources to address the issue of “making Western long-term historical statistics databases work for us”; at the same time, through “top-level design” and “unified layout”, we should attempt to construct domestic and global long-term historical statistics databases.
[The notes of this article are omitted]