Abstract
Many phenomena of high relevance for economic development such as human capital, geography and climate vary considerably within countries as well as between them. Yet, global data sets of economic output are typically available at the national level only, thereby limiting the accuracy and precision of insights gained through empirical analyses. Recent work has used interpolation and downscaling to yield estimates of sub-national economic output at a global scale, but respective data sets based on official, reported values only are lacking. We here present DOSE — the MCC-PIK Database Of Sub-national Economic Output. DOSE contains harmonised data on reported economic output from 1,661 sub-national regions across 83 countries from 1960 to 2020. To avoid interpolation, values are assembled from numerous statistical agencies, yearbooks and the literature and harmonised for both aggregate and sectoral output. Moreover, we provide temporally- and spatially-consistent data for regional boundaries, enabling matching with geo-spatial data such as climate observations. DOSE provides the opportunity for detailed analyses of economic development at the subnational level, consistent with reported values.
Similar content being viewed by others
Background & Summary
Understanding what drives or impedes economic progress is of key importance to various aspects of global welfare such as the alleviation of poverty and reduction of inequality. Previous research has highlighted e.g. the role of institutions1,2,3, human capital4,5, geography6,7,8,9 and climate10,11,12,13,14. These factors can differ substantially not only across but also within countries. Yet, development indicators such as Gross Domestic Product (GDP) are typically available at the national level only (or provided at higher spatial detail for individual countries only), therefore limiting global analyses to the level of entire countries and hindering the detail and potential accuracy of their insights15. As such, there is a clear need for more spatially-detailed economic data on a global scale.
A few attempts to generate such data have already been made. One of the earliest versions of a global, sub-national economic data set comes from Nordhaus6, who developed a 1° × 1°-grid of per capita economic output to better understand the relationship between geography and economic activity. While innovative, this data set relies heavily on interpolation using population census data. In 2013, Gennaioli et al.4 published the first consistent sub-national economic data with administrative regions (as opposed to a grid), but at five year increments and partly also relying on interpolation across time. Building on the work of Gennaioli et al., Kummu et al.16 produced a “gap-filled” global gridded data set which also included the Human Development Index scores as well as GDP per capita and total GDP for all years between 1990 and 2015. Here, significant interpolation was necessary to fill gaps across time and space. Perhaps the closest example to the here presented DOSE data set17 in terms of spatial and temporal coverage is the Subnational Human Development Database by Smits and Permanyer18, which provides regional income per capita (amongst other indicators) for 1625 regions across 161 countries for the period 1990 to 2017. Unlike DOSE17, however, this data set relies on household surveys and interpolation between survey years for the majority of non-OECD countries. Other examples of sub-national GDP data include a gridded data set for China19 as well as another global grid (1 km × 1 km) constructed using nighttime light data as a proxy for GDP20. Although providing global coverage with consistent methodology, the use of nighttime light data as proxy for economic activity is subject to limitations. In particular, economic activities that do not necessarily emit additional nighttime light as they grow such as agriculture and forestry are not well captured if no additional data proxies are used21. This can lead to an underestimation of economic activity in rural, darker areas whilst overestimating it in areas with exceptionally high light22,23.
Sub-national economic data have been used in a variety of research related to regional development. For instance, Gennaioli et al.24 compared within- and across-country economic development, particularly focusing on the rate of income convergence. Cuaresma et al.25 assessed the determinants of economic growth in European regions. A third study, conducted by Lee et al.26, explored the connection between sub-national government institutions and economic development in 340 African regions, concluding a significant, positive relationship.
Despite these developments in the production and use of sub-national data, their suitability for global empirical analyses is still challenged by the use of interpolation across time and space. Whereas spatial interpolation is problematic for cross-sectional analyses, temporal interpolation is particularly of concern in the context of time series and panel regression analyses that exploit short-run variations in data, which would naturally be obscured by the use of interpolation spanning multiple years. In order to make accurate inferences about environmental and social phenomena that operate at shorter timescales, there is hence a need for non-interpolated data that better mirrors natural variation as it occurs.
To address these challenges, we present DOSE — the MCC-PIK Database Of Sub-National Economic output17. DOSE is a global data set in which aggregate and sectoral economic data from over 1,660 regions across the world have been harmonized without interpolation. Instead, DOSE17 builds on a unique collection of data from various data sources, including statistical agencies, yearbooks and academic literature - all of which have been recorded in Table 1. In total, the data span 83 countries with annual temporal coverage from 1960 to 2020 (although temporal coverage varies for different regions). This yields more than 46,000 observations, each containing economic output for a particular region in a given year. At minimum, each observation contains the total gross regional product (GRP). In the majority of observations, economic output data are also provided for three main sectors: agriculture, manufacturing, and services (referred to as sectoral GRP). In addition to regional economic output in local currency units (LCU) at current market prices as generally collected from the original data sources, we provide per capita estimates in LCU and US dollars at both, current and 2015 market prices to enable comparison across time and space. Importantly, unlike other similar data sets, DOSE17 does not use inter- or extrapolation across time to estimate values between or outside of the periods with real observations. Instead, it contains output values that reflect the figures reported by the raw data sources. Despite this, we are still able to provide data on an annual basis rather than incrementally, considerably improving the coverage compared to previous databases4. This annual resolution makes DOSE17 suitable not only for analysis of long-run economic development but also of short-run variations in output data. This is a distinct advantage that DOSE17 holds over previous iterations of global sub-national economic data sets, where short-term inter-temporal variation is often obscured. Furthermore, the economic data in DOSE17 reflect values in sub-national regions, avoiding the spatial interpolation used to produce gridded data sets6,16. The non-interpolated, sub-national detail in DOSE17 allows for a variety of potential research directions, with one prominent example being climate impact analyses. Weather and climate are rarely experienced homogeneously at the national level27. Consequently, the impacts of weather extremes on production may not be identified precisely by means of national-level data. DOSE17 can help better understand the heterogeneity in the resulting economic impacts. It is also suited for exploring more general political-economic questions of within-country development, or for case studies related to a specific economy or set of economic regions. Due to the large number of observations across both time and space, DOSE17 allows for robust cross-sectional and inter-temporal analyses (for example via fixed-effects panel regression).
A schematic overview of the processes involved in the creation of DOSE17 is presented in Fig. 1 and a detailed description can be found in the Methods section. Annual, sub-national economic data were collected from various sources like EuroStat, the OECD database, national statistical institutes, and academic literature. Population data stem from the same data sources as the economic data, exchange rates from the FRED database and national price indices from the World Bank. The latter two were used to provide estimates that account for currency differences and inflation, respectively. We also provide geo-spatial data describing the administrative boundaries of the regions in DOSE17. These consider shifting administrative boundaries over time and can be used to flexibly combine DOSE17 with any other geo-referenced data set, for example to study the sub-national economic impacts of climate change. An earlier version of the DOSE data set has already been used in several of such impact-related studies, for example to assess the role of temperature and precipitation averages, variability and extremes in affecting economic growth12,13,14,28. Compared to this earlier version, we have here substantially updated and extended DOSE17, both in space and time as well as with respect to the inclusion of further variables reflecting different conversion and deflation methods.
In the following sections, we present a detailed account of the steps and methods involved in creating DOSE17, the data records it produced, a documentation of our technical validation, additional usage notes for those interested in its application, as well as access links.
Methods
Collection of raw economic data (M1)
The first step in the creation of DOSE17 consisted of the systematic collection of raw economic output data at the sub-national level. This step is represented schematically by the top left corner of Fig. 1, labeled M1. We used the first administrative division below national for consistency, corresponding to the state-level in the USA, Bundesländer-level in Germany, or equivalent. For countries in the European Union, the EuroStat database was used to obtain Gross Regional Product (GRP, total economic output) and Gross Value Added (GVA) by industry (sector-level output) data for all NUTS2 level regions. For OECD and some non-OECD countries, GRP data stem from the OECD Regional Database for TL2-regions. For the remaining countries, we used their respective national statistical databases, annual reports, or reviewed the literature to collect GRP and GVA by industry data. For some countries, GRP data were not available. Instead, we used the combined GVA of all industries as an estimate of GRP. A few other countries provided GRP by industry instead of GVA. If both total GRP and GVA by industry were not available, we used (household) income data as a proxy. This was the case for Ireland, Kenya, Pakistan, and Paraguay. For these countries, additional data such as average household size was necessary to calculate an estimate for total GRP.
In most of the raw data, sector-level GVA data (or GRP if GVA was not available) were pre-classified into economic activities based on the International Standard Industrial Classification of All Economic Activities (ISIC). Within DOSE17, these activities were further aggregated into three main sectors: agriculture, manufacturing, and services, as detailed in Table 2. Total GRP was generally collected separately, constituting the main estimate of aggregate GRP in DOSE17. Minor discrepancies between this total GRP value and the sum of the three sectoral outputs exist due to inconsistencies in the original national accounting. When estimating sectoral shares users are therefore advised to use the sum of output from the three sectors rather than the direct estimate of total GRP. In some cases, total GVA was used rather than total GRP due to a lack of data availability.
If the raw data source differed for a given country over time or the respective institute provided a different structure of GVA by industry, we ensured to aggregate over similar categories to have consistent sectoral coverage over time. With Ecuador, for example, the original raw data were split into economic activities that were more specific than those in the ISIC classification. Agricultural output in particular was separated into eight categories based on crop and animal type. In such cases, we made common-sense judgements to determine to which of the three sectors in DOSE17 these more specific activities should be allocated. For some countries, sectoral data were unavailable.
In total, we were able to assemble economic output data for 1,661 regions in 83 countries around the globe. The full spatial coverage is represented on the world map in Fig. 2. Coverage of GRP data over time is shown in Fig. 3. As can be seen in Fig. 2, temporal coverage varies across regions. There are comparably long time series with more than 40 observations per region across Northern America, Europe, East Asia, Oceania, Latin-America and India. By contrast, data were only available from 1990 or later on-wards for Russia and countries of the former Soviet Union. Data coverage is particularly sparse for African and Middle-Eastern countries where official reports of sub-national economic data are largely lacking. Kenya constitutes an exception here as data on labor earnings (1969–2012) and GRP (2013–2019) are available from 1969 onwards. Furthermore, we were also able to collect sub-national economic data for more than 20 years for several South-east African countries (South Africa, Mozambique, and Tanzania). In comparison, data for North African countries were available for less than ten years per region. Table 1 provides an overview of all raw data sources and Table S1 in the supplementary material a detailed list of raw data per country. In the cases that raw data sources changed over time, we document these structural changes with a categorical variable “StructChange”, which e.g. takes the value of 1 in those years where a new raw data source has been introduced for the first time. Raw data were generally collected at current market prices in local currency.
Processing of economic data (M2)
In order to compare economic data across regions, we calculated per capita figures, using sub-national population estimates from the same data sources as the economic output. Total and sectoral GRP data were divided by the regional population to obtain per capita output in current market prices. We include the population estimates in the final data set to facilitate conversion between per capita and absolute figures.
An overview of all auxiliary data used and variables constructed in the processing step can be found on the right side of Fig. 1, labeled M2. All variables contained in DOSE17 are listed in Table 3. In addition to nominal per-capita sectoral and gross regional products in local currency units, we also provide estimates in constant 2015 LCU values as well as in current and constant US dollars (along with the data used for these conversions), thus enabling the use of our data in a broad range of contexts and applications.
Two types of data were used in these conversions, namely, market exchange rates and national-level GDP deflators. Both indices consistently account for historic currency reforms when necessary. Market exchange rate data were taken from the FRED database of the Federal Reserve Bank of St. Louis, specifically the “Penn World Table 10.0”29. The specific format of these data is national currency units per US dollar, collected annually. Using these annual exchange rates, we converted sectoral and total GRP in local currency per capita to US dollar equivalents. An alternative option is to use a Purchasing Power Parity (PPP) exchange rate as it tends to better reflect price differences in consumer goods across countries. Since non-tradeable goods and services are cheaper in lower-income countries, the use of market exchange rates can underestimate the purchasing level of incomes in these countries. Another limitation of using market exchange rates to convert to a common currency is that under political influence, they may not accurately reflect a nation’s real economy. For example, in countries that have pegged their currency to the dollar, the market exchange rate is subject to government regulations that could misrepresent true purchasing power. In such contexts, PPP exchange rates may offer a more accurate conversion. However, due to the interpolation of PPP between survey years which occur in approximately 6-year increments30, we opted for market exchange rates to avoid introducing an interpolation bias which would pose particular problems for time-series and panel analyses. In the context of cross-sectional analyses, by contrast, the use of PPP-based values may be favorable. For example, when visualising the distribution of incomes across regions in Figs. 4, 5, as well as when comparing our data to another database which is based on PPP values (Fig. 7), we show PPP-adjusted values, i.e. for the generation of these values PPP instead of market exchange rates were used. We hence also provide PPP exchange rates within the DOSE data set17 to enable PPP-based conversions.
In order to convert current prices into constant 2015 prices, we used national GDP deflator data from the World Bank. By using one global data source, we minimize the risk of country-specific measurement irregularities. As the World Bank deflator data set originally used different base years for each country, we harmonized the deflator time series to base year 2015. We chose to use a GDP deflator because unlike a consumer price index (CPI) it also takes into account price changes in purchases made by businesses, governments, and foreigners in addition to those made by regular consumers. Furthermore, a GDP deflator ignores price changes from importers. Unfortunately, deflator data are generally unavailable at the sub-national level, such that regional variations in prices are not accounted for.
In general, there are two approaches to make economic data comparable between regions and across time using market exchange rates and national deflator data, both of which we produced with our data. The approaches differ in the order with which they deal with the inter-national and inter-temporal components of currency comparison and each have their strengths and shortcomings. The first approach given in Eq. 1 converts current per-capita GRP of region r in year y in local currencies, grp_pc_lcur, y, first into current US dollars, using the market exchange rate fxr,y for year y, and then applies the US deflator with base year 2015, deflator_2015USA,y, to obtain values at constant prices.
With such an approach the lack of sub-national price indices is bypassed, but under the assumption that goods are perfectly tradeable between countries. Similar procedures for sub-national entities have already been applied in some contexts. For instance Kalkuhl & Wenz12 used GRP in current US dollars and applied region-specific polynomial time trends and year-fixed-effects to control for changes in regional prices.
A second approach, given in Eq. 2, controls first for local nominal price changes prior to applying the market exchange rates for the year 2015.
In particular, the local national deflator, deflator_2015r,y, is used to calculate GRP in constant 2015 prices, after which the US dollar exchange rate in 2015, fxr,2015, is applied to convert to constant US dollars. The advantage of the second approach lies in its correction of national inflationary tendencies that are not well represented by market exchange rates. However, the application of national deflators as the main correction for inter-temporal price changes may not well reflect differential within-country price changes which emerge over time31. The use of sub-national price indices could mitigate this issue, but such data are not currently available for the majority of regions.
We here provide estimates resultant from both methods as well as nominal values in either LCU or US dollars such that users can employ the variable that is most suitable in their context. For the same reason, the DOSE data set17 also contains Worldbank’s CPI data with base year 2015. In the section on technical validation, we compare both conversion methods to a previous sub-national database24 which instead used sub-national GRP in constant local prices to downscale national GDP in constant, US dollars. We find that both methods compare similarly well, with slightly greater consistency when using our first method for accounting for currency and inflation changes.
The resulting economic variables can be used for a variety of analyses. We provide a few basic examples here. Figure 4 shows how GDP and sectoral output were distributed across the world in 2013. In Fig. 5 one can observe more detailed within-country differences for the example of Europe. Figure 6 demonstrates the possibility to examine trends in economic growth over time, here exemplified for Chinese provinces.
Matching DOSE with spatial data (M3)
To accompany the economic data in DOSE17, we provide unique identifiers for each region that can be used to match them to the relevant administrative boundaries (see section M3 in Fig. 1). Specifically, we provide two identifiers for each region: one for the first subnational level boundary, GID_1, and a second for the national level it belongs to, GID_0. Users can find spatial data at the same regional level as our economic data (first sub-national level) for all but eight countries by using the Database of Global Administrative Areas (GADM, https://gadm.org/download_country_v3.html), in particular version 3.6.
In the eight exceptional countries, the regions listed in a given country’s economic data did not match with any of the regions provided by GADM’s first sub-national level. This was due to one of two reasons: (i) the economic data were collected at a level that was either finer or coarser than any of the administrative levels covered by GADM level 1, or (ii) region names and administrative boundaries had changed either over time or between the time that the economic data and GADM version 3.6 were published. In such cases, we provide the spatial data ourselves. Shape-files were gathered from alternative sources, like the Humanitarian Data Exchange, or created manually (i.e., by aggregating smaller administrative boundaries). For example, in Macedonia, GADM level 1 lists 85 sub-national regions whereas we collected economic data for only eight regions. Here we aggregated the smaller regions from GADM level 1 to match the larger regions for which economic data were available (using QGIS and sources such as http://citypopulation.de/en/northmacedonia/cities/). Although this division does not reflect the official administrative division of the first sub-national level but rather the division of statistical regions, we preferred to provide a broader regional division instead of none. We then created GID_1 identifiers for these ‘custom’ regions (starting from MKD.100_1, MKD.101_1, etc. to ensure that there was no overlap with the official GADM level 1 identifiers, which go from 1–85).
In rare cases, the set of sub-national regions in a country’s economic data changed over time. In the case of Morocco, for instance, the set of regions in the economic data for 2007–2013 differs from those for 2014–2019. To preserve this distinction, we treat the two sets of regions as separate entities, providing spatial data for the regions not found in GADM 3.6, and unique identifiers for both sets of regions. In Morocco’s case, GADM level 1 data could be matched to the regions for 2014–2019, whilst the Humanitarian Data Exchange was used to acquire spatial data for the 2007–2013 regions, which were then given a unique GID_1 identifier.
Table 4 shows the full list of countries whose economic regions did not initially match with the GADM level 1 division or any other level. Spatial data from GADM level 1 can be merged with the spatial data from Table 4 regions to produce a comprehensive shapefile with geometry for each region in DOSE17. The extra procedures applied to the eight exceptional countries (e.g. Macedonia and Morocco) ensure spatial and temporal consistency with the observations in DOSE17.
Matching climatic data
Assessments of economic climate impacts are a typical example of using DOSE17 in conjunction with detailed geo-spatial data12,13,14. Within the data set, we provide estimates of sub-national annual mean temperature and annual total precipitation having aggregated high-resolution climate data from the ERA-5 reanalysis32 to the administrative boundaries described in the previous section (see Matching DOSE with Spatial Data). We calculate annual mean temperature and annual total precipitation from daily 2-m air temperatures and daily precipitation totals at the grid-cell level (0.25-by-0.25 degrees), before aggregating to the sub-national administrative boundaries using an area-weighted average. In case a grid cell is split between two regions or a region and the ocean, we assess the proportion of the grid-cell that falls within the administrative boundary of a region and use this to weight its contribution to the sub-national average of that region.
A sample visualization using these climate variables can be found in Figure S1 in the Supplementary Information.
Data Records
The data set and corresponding spatial data are available on Zenodo at the following repository https://doi.org/10.5281/zenodo.757324917. The main file DOSE_V2.csv contains the raw economic data, typically gross regional product at current market prices in local currencies, grp_lcu, and population data, pop, as collected from the different data sources for 1,661 sub-national regions in 83 countries from 1960–2020. For Australia, regional GRP data are also available and included for 1953. If there was a change in either the data source or the administrative boundaries of a region, this is indicated by the StructChange column. The main data file further comprises the processed economic data, i.e. per-capita values at current and constant prices in local currencies and USD, as described in the Methods section (grp_pc_lcu, grp_pc_lcu_2015, grp_pc_lcu2015_usd, grp_pc_usd and grp_pc_usd_2015). Deflator data, deflator_2015, and exchange rates, fx, used to generate these different economic variables are given as well. GRP data are also provided at the sector-level for the agriculture, manufacturing/industry and services sectors, indicated by the prefixes ag_, man_ and serv_, respectively. To facilitate further analyses, 2015 Consumer Price Indices, cpi_2015, and Purchasing Power Parities, PPP, as well as annual temperature, T_a, and precipitation, P_a, data are included. For each country and region, there are unique identifiers, GID_0 and GID_1, given. Each column of the main file DOSE_V2.csv is described in full in Table 3.
The second file, DOSE_shapefiles.gpkg, is a composite shapefile (packaged in the geopackage format) that contains administrative boundaries for the eight exceptional countries listed in Table 4 for which spatial data cannot be obtained from GADM version3.6, along with the region names and unique identifiers that can be used to match them with DOSE_V2.csv.
Technical Validation
The technical validation of DOSE17 was carried out in two distinct processes: (1) manual validation and (2) comparison with similar, pre-existing data. Each of these steps is explained in detail in the subsequent sections, followed by a brief discussion of the data set’s limitations.
Manual validation
To ensure data quality, we manually inspected the timeseries throughout the data collection process. This involved checking for missing values, outliers, or unexpected patterns in the data, as well as identifying any discrepancies or errors that needed to be corrected. When collecting economic and population data at the sub-national level, we ensured the sum of all regions resembled the indicator’s value at the national level. We also examined possible differences between the sum of the computed sectoral GRP values and their total GRP value in a given year, which may occur because of different considerations of financial intermediaries or product taxes by the statistical institute. We determined that there were no instances where this gap was unreasonably large, but researchers are advised to compute sectoral shares based on the sum of the three sectors instead of using total GRP as base value. We validated the time series’ consistency by closely inspecting any jumps in the data. If one occurred, we compared it with data from other regions in the same country. If it appeared to be a year of economic slowdown for the whole country, we compared our data with national GDP records provided by other sources, such as World Bank, to exclude any error in the raw data. For example, Egyptian regions showed a strong decrease in economic production between 2000 and 2003 which we initially flagged as questionable, but could then confirm by comparing it to the national GDP values reported by the World Bank. Irregularities also occurred when the raw data were available in PDF format only (e. g. Kenya, Vietnam). For very large data sets, we made use of software that converted these data to spreadsheet format. In these cases, validating the conversion’s accuracy retrospectively was of high relevance. A common mistake of the software was adding an additional 0 which increased the data by one decimal power. Comparing the indicators to those at the national level revealed these cases reliably.
After having collected and merged the sub-national data, we used the coefficient of variation (CV), i.e. the ratio of the standard deviation to the mean, in output data for each region over time as a simple metric to further identify potential outliers (see Figure S2 in the Supplementary Information). In most cases, coefficients of variation are far below 1, with a few exceptions. Particularly large values were observed in China (CV∈[1, 1.5]), but these were found to reflect the exceptionally long time periods of large economic growth rates since the 1960s rather than inconsistencies in the raw data. Similarly large values for Kenya were also identified (CV∈1,2), and in this case found to arise due to the change in data sources from labor earnings data to GRP between 1999 and 2013, changes which are accounted for by our use of categorical variables indicating structural changes (see Methods section “M1. Raw Economic Data Collection”).
Comparison to Pre-existing data sources
As outlined in the background section, some data sets of sub-national economic output are already available, offering the opportunity to compare these data to DOSE17 for a general validation of our estimates. The most comparable source is that provided by Gennaioli et al.24, which constitutes values for sub-national entities (rather than gridded values) but only at five year increments, for some regions beginning in 1950. However, Gennaioli et al.24 (referred to as G2014 from hereon) used a different approach to account for inflation and currency exchanges. They first estimated the share which each region contributes to the national sum of regional GRP in current, local prices. This share was then used to downscale national GDP in constant, PPP international dollars to the sub-national level at five year increments. Since we have used different approaches to address inflation and currencies as discussed in the Methods section, neither data set should be considered the “ground-truth”, but the comparison should nevertheless facilitate confidence in the general level of consistency. To make our data as comparable as possible to G2014, we use PPP-exchange rates rather than market ones. We compared the indicator of G2014 to both of our PPP-adjusted variables (one for each conversion technique outlined in the section “Processing of Economic Data”).
We identify approximately 5,600 observations of a total of 9,500 in G2014 which are comparable across both data sets, shown in Fig. 7 as scatterplots. The left four panels (a, c, e, and f) are based on the PPP-adjusted values estimated using the first method outlined in the section “Processing of Economic Data” (first converting to a common currency, then applying a US deflator, see eq. (1)). Conversely, the right hand panels (b, d, f and h) use the second conversion technique (first applying a national deflator, then converting to a common currency). In panels c and d, the comparison and correlation is made after both indicators have been demeaned by the national average GRP in each year, to emphasise consistencies and discrepancies between the two data sets in the within-country dimension. In panels e and f, this comparison is made after a demeaning by the national GRP value in all years to emphasise both the within-country and temporal dimension. In panels g and h, values are demeaned by the average GRP of each subnational region across all years to emphasise the temporal dimension only.
The comparison across countries (Fig. 7a,b, see Figure S3 in the Supplementary Information for a comparison on a logarithmic-scale) shows a high level of similarity and correlation (correlation coefficient, r > 0.88) for both conversion methods. High similarities (r > 0.7) are still obtained when comparing the data having been demeaned by country-year, country or sub-national region in order to focus on the within country and temporal dimensions (Fig. 7c–h). With regards to the different methods for currency conversion outlined in section “Economic Data Processing”, method 1 appears to provide generally greater consistency to the down-scaling method of G2014, particularly along the temporal dimension. When using non PPP-adjusted values to make this comparison, similarly high correlations are observed, with generally better correlations using the currency conversion order of method 2 (Figure S4 in the Supplementary Information). Overall, these comparisons validate the consistency of DOSE17 with previous estimates of sub-national output along multiple dimensions, strengthening confidence in our methodology and the additional 35,000 observations which DOSE17 provides in addition to theise previous data sets24.
Limitations
There are a few limitations of DOSE17 that users should be aware of. Despite the increased spatial and temporal coverage of DOSE17 in comparison to most pre-existing datasets, data gaps in both dimensions remain. For example, we are still missing sub-national output data for a large number of African and Middle-Eastern countries (compare Fig. 2). This implies that regions in relatively wealthier countries are over-represented whereas poorer ones are under-represented. Temporally, our panel data also tends to be unbalanced with the majority of observations taking place over the last three decades of coverage (1990–2020) compared to the earlier decades (1960–1990). With respect to future developments, the use of satellite-derived data products could be a promising avenue for filling in spatial gaps.
Another limitation is that converting sub-national nominal GRP values in local currencies to real GRP data in USD is not straight-forward, partly due to the lack of auxiliary data at the sub-national level (such as GDP deflators which are generally unavailable at the global scale). Consequently, we provide a range of auxiliary data, allowing for different conversion procedures as described in detail in the Processing of Economic data (M2) section.
Finally, we acknowledge that the accuracy of DOSE17 is inherently limited by the accuracy and authenticity of national and regional administrations that report their economic output. The ‘reported’ nature of the values in DOSE17 should be emphasized in any future endeavors that exploit the data. It should be noted that some administrations may be incentivized to overestimate yearly GDP output, particularly in autocratic systems of governance33. Bearing this in mind, we believe there is great value in global databases of reported figures. Amongst other uses, they themselves form the basis of assessments of the integrity of reported figures and the integration of new observational methodologies such as nighttime light data.
Usage Notes
Because of its.csv-format, the usage of DOSE is straight-forward. Researchers should be aware that when calculating sectoral shares of GRP, these should be assessed with respect to the sum of the three sectoral outputs. As the sectoral output data are generated by aggregating over specific categories of Gross Value Added by industry, GRP and the sum over the three sectors may not always equal one another. Regarding the different inflation and currency conversions, users can choose that economic variable which is most suitable for their research interest. The custom shapefile are provided in a.gpkg (geopackage).
Code availability
Code can be found in a public repository at: https://doi.org/10.5281/zenodo.765959934. This includes code to replicate the figures shown here as well as code to merge spatial data from GADM level 1 with the spatial data provided for eight exceptional countries as presented in Table 4 to produce a comprehensive shapefile with geometry for each region in DOSE.
References
Siddiqui, D. & Ahmed, Q. The effect of institutions on economic growth: A global analysis based on gmm dynamic panel estimation. Structural Change and Economic Dynamics 24, 18–33, https://doi.org/10.1016/j.strueco.2012.12.001 (2013).
Acemoglu, D., Johnson, S. & Robinson, J. A. Institutions as a fundamental cause of long-run growth. In Aghion, P. & Durlauf, S. N. (eds.) Handbook of economic growth, vol. 1 of Handbook of Economic Growth, chap. 6, 385–472, https://doi.org/10.1016/S1574-0684(05)01006-3 (Elsevier, 2005).
Acemoglu, D., Gallego, F. A. & Robinson, J. A. Institutions, human capital, and development. Annual Review of Economics 6, 875–912, https://doi.org/10.1146/annurev-economics-080213-041119 (2014).
Gennaioli, N. & La Porta, R. Lopez-de Silanes, F. & Shleifer, A. Human capital and regional development. The Quarterly Journal of Economics 128, 105–164, https://doi.org/10.1093/qje/qjs050 (2013).
Barro, R. J. Economic Growth in a Cross Section of Countries*. The Quarterly Journal of Economics 106, 407–443, https://doi.org/10.2307/2937943 (1991).
Nordhaus, W. D. Geography and macroeconomics: New data and new findings. Proceedings of the National Academy of Sciences 103, 3510–3517, https://doi.org/10.1073/pnas.0509842103 (2006).
Sachs, J. D. & Warner, A. M. Fundamental sources of long-run growth. American Economic Review 87, 184–188 (1997).
Gallup, J. L., Sachs, J. D. & Mellinger, A. D. Geography and economic development. International Regional Science Review 22, 179–232, https://doi.org/10.1177/016001799761012334 (1999).
Krugman, P. The role of geography in development. International Regional Science Review 22, 142–161, https://doi.org/10.1177/016001799761012307 (1999).
Dell, M., Jones, B. F. & Olken, B. A. Temperature shocks and economic growth: Evidence from the last half century. American Economic Journal: Macroeconomics 4, 66–95, https://doi.org/10.1257/mac.4.3.66 (2012).
Burke, M., Hsiang, S. M. & Miguel, E. Global non-linear effect of temperature on economic production. Nature 527, 235–239, https://doi.org/10.1038/nature15725 (2015).
Kalkuhl, M. & Wenz, L. The impact of climate conditions on economic production. evidence from a global panel of regions. Journal of Environmental Economics and Management 103, 102360, https://doi.org/10.1016/j.jeem.2020.102360 (2020).
Kotz, M., Wenz, L., Stechemesser, A., Kalkuhl, M. & Levermann, A. Day-to-day temperature variability reduces economic growth. Nature Climate Change 11, 319–325, https://doi.org/10.1038/s41558-020-00985-5 (2021).
Kotz, M., Levermann, A. & Wenz, L. The effect of rainfall changes on economic production. Nature 601, 223–227, https://doi.org/10.1038/s41586-021-04283-8 (2022).
Henderson, J. V., Storeygard, A. & Weil, D. N. Measuring economic growth from outer space. American Economic Review 102, 994–1028, https://doi.org/10.1257/aer.102.2.994 (2012).
Kummu, M., Taka, M. & Guillaume, J. H. Gridded global datasets for gross domestic product and human development index over 1990–2015. Scientific Data 5, 1–15, https://doi.org/10.1038/sdata.2018.4 (2018).
Wenz, L., Carr, R., Kögel, N., Kotz, M. & Kalkuhl, M. DOSE - Global dataset of reported subnational economic output. Zenodo https://doi.org/10.5281/zenodo.7573249 (2023).
Smits, J. & Permanyer, I. The subnational human development database. Scientific Data 6, 190038, https://doi.org/10.1038/sdata.2019.38 (2019).
Dong, L., Yuan, X., Li, M., Ratti, C. & Liu, Y. A gridded establishment dataset as a proxy for economic activity in china. Scientific Data 8, 1–9, https://doi.org/10.1038/s41597-020-00792-9 (2021).
Chen, J. et al. Global 1 km × 1 km gridded revised real gross domestic product and electricity consumption during 1992–2019 based on calibrated nighttime light data. Scientific Data 9, 1–14, https://doi.org/10.1038/s41597-022-01322-5 (2022).
Keola, S., Andersson, M. & Hall, O. Monitoring economic development from space: Using nighttime light and land cover data to measure economic growth. World Development 66, 322–334, https://doi.org/10.1016/j.worlddev.2014.08.017 (2015).
Gibson, J., Olivia, S. & Boe-Gibson, G. Night lights in economics: Sources and uses 1. Journal of Economic Surveys 34, 955–980, https://doi.org/10.1111/joes.12387 (2020).
Wang, T. & Sun, F. Global gridded gdp data set consistent with the shared socioeconomic pathways. Scientific data 9, 1–10, https://doi.org/10.1038/s41597-022-01300-x (2022).
Gennaioli, N., La Porta, R., Lopez De Silanes, F. & Shleifer, A. Growth in regions. Journal of Economic Growth 19, 259–309, https://doi.org/10.1007/s10887-014-9105-9 (2014).
Cuaresma, J. C., Doppelhofer, G. & Feldkircher, M. The determinants of economic growth in European regions. Regional Studies 48, 44–67, https://doi.org/10.1080/00343404.2012.678824 (2014).
Iddawela, Y., Lee, N. & Rodríguez-Pose, A. Quality of sub-national government and regional development in Africa. The Journal of Development Studies 57, 1282–1302, https://doi.org/10.1080/00220388.2021.1873286 (2021).
IPCC. Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, Cambridge, UK and New York, NY, USA, 2022).
Callahan, C. W. & Mankin, J. S. Globally unequal effect of extreme heat on economic growth. Science Advances 8, https://doi.org/10.1126/sciadv.add3726 (2022).
Feenstra, R. C. & Inklaar, R. Penn world table 10.0, https://doi.org/10.15141/S5Q94M (2021).
World Bank. How do you extrapolate the PPP conversion factors estimated by the ICP? – World Bank Data Help Desk. Available at https://datahelpdesk.worldbank.org/knowledgebase/articles/665452-how-do-you-extrapolate-the-ppp-conversion-factors.
Blanchet, T. Prices and currency conversions in WID. world. WID.world Technical Note Series 3, 16 (2017).
Hersbach, H. et al. The era5 global reanalysis. Quarterly Journal of the Royal Meteorological Society 146, 1999–2049 (2020).
Martínez, L. R. How Much Should We Trust the Dictator’s GDP Growth Estimates? Journal of Political Economy 130, 2731–2769, https://doi.org/10.1086/720458. Publisher: The University of Chicago Press (2022).
Wenz, L., Carr, R. D., Koegel, N., Kotz, M. & Kalkuhl, M. Code for the publication “DOSE - Global data set of reported sub-national economic output” https://doi.org/10.5281/zenodo.7659600 (2023).
Eurostat. Gross domestic product (GDP) at current market prices by NUTS 2 regions. Available at https://ec.europa.eu/eurostat/databrowser/view/nama_10r_2gdp/default/table?lang=en.
Eurostat. Gross value added at basic prices by NUTS 3 regions. Available at https://ec.europa.eu/eurostat/databrowser/view/NAMA_10R_3GVA/default/table?lang=en.
Eurostat. Regional yearbooks – various editions. https://doi.org/10.2785/915176.
OECD. Regional economy. https://doi.org/10.1787/6b288ab8-en (2023).
Instituti i Statistikave - Albania National Statistics. Gross Domestic Product by Statistical Regions NUTS (NACE Rev.2) by Region, Type and Years. Available at http://databaza.instat.gov.al/pxweb/en/DST/START__NA__NAY2/NAR_02/.
National Statistical Institute of Argentina. INDEC: Instituto Nacional de Estadística y Censos de la República Argentina. Available at https://www.indec.gob.ar/indec/web/Institucional-Indec-SistemaEstadistico.
Australian Bureau of Statistics. Australian National Accounts: State Accounts. Available at https://www.abs.gov.au/statistics/economy/national-accounts/australian-national-accounts-state-accounts.
Statistics Austria. Regional accounts. Available at https://www.statistik.at/en/statistics/national-economy-and-public-finance/national-accounts/regional-accounts.
The State Statistical Committee of the Republic of Azerbaijan, A. System of national accounts and balance of payments. Available at //www.stat.gov.az/source/system_nat_accounts/?lang=en.
Government of the Bahamas - Department of Statistics. National Accounts Report 2020. Available at https://bahamas.gov.bs/.
National Statistical Comittee of the Republic of Belarus. National accounts of the Republic of Belarus, Available at https://www.belstat.gov.by/en/ofitsialnaya-statistika/publications/statistical-publications-data-books-bulletins/public_compilation/index_27957/ (2021).
Bolivia - Instituto Nacional de Estadisticas INE. Producto Interno Bruto Departamental. Available at https://www.ine.gob.bo/index.php/estadisticas-economicas/pib-y-cuentas-nacionales/producto-interno-bruto-departamental/producto-interno-bruto-departamental/.
Agency for Statistics of Bosnia and Herzegovina. National Accounts. Available at https://bhas.gov.ba/Calendar/Category?id=12&page=4&statGroup=12&tabId=0&lang=en.
Instituto Brasileiro de Geografia e Estatística. Regional Accounts of Brazil | IBGE. Available at https://www.ibge.gov.br/en/statistics/economic/national-accounts/16855-regional-accounts-of-brazil.html?edicao=35525&t=resultados.
National Statistical Institute of Bulgaria. GDP and GVA by Economic Sector and Region. Available at https://infostat.nsi.bg/infostat/pages/reports/query.jsf?x_2=1262.
Statistics Canada - Government of Canada. Table 36-10-0402-01 Gross domestic product (GDP) at basic prices, by industry, provinces and territories. Available at https://doi.org/10.25318/3610040201-eng.
Banco Central Chile. Gross domestic product by region, current prices, linked series, reference 2018 (billions of pesos). Available at https://si3.bcentral.cl/Siete/EN/Siete/Cuadro/CAP_CCNN/MN_CCNN76/CCNN2018_PIB_REGIONAL_N/637899740344107786.
National Bureau of Statistics of China. National Data - Annual by Province. Available at https://data.stats.gov.cn/english/easyquery.htm?cn=E0103.
Colombia National Adminstrative Department of Statistics - DANE. Departmental National Accounts - Results by Economic Activity. Available at https://www.dane.gov.co/index.php/en/statistics-by-topic-1/national-accounts/departmental-national-accounts.
Croatian Bureau of Statistics. Gross Domestic Product - Reveiw by Counties. Available at https://podaci.dzs.hr/media/rihnbma3/bruto-domaci-proizvod.xlsx.
Czech Statistical Office. Database of Regional Accounts. Available at https://apl.czso.cz/pll/rocenka/rocenkavyber.volba?titul=Selected%20indicators%20by%20industry&mypriznak=RD&typ=1&proc=rocenka.presB&mylang=EN&jak=4.
Statistics Denmark. Production and generation of income (10a3-grouping) by region, transaction, industry and price unit. Available at https://www.statbank.dk/10021.
Banco Central del Ecuador. Cuentas Provinciales (anuales). Available at https://contenido.bce.fin.ec/documentos/Estadisticas/SectorReal/CuentasProvinciales/Indice.htm.
Egyptian Ministry of Planning and Economic Development. GDP by Governorate. Available at https://mped.gov.eg/Governorate?lang=en.
Statistics Estonia. RAA0051 GROSS DOMESTIC PRODUCT BY COUNTY AND ECONOMIC SECTOR. Available at https://andmed.stat.ee:443/pxweb/en/stat/stat__majandus__rahvamajanduse-arvepidamine__sisemajanduse-koguprodukt-(skp)__regionaalne-sisemajanduse-koguprodukt/RAA0051.px/ (ESA 2010).
Bureau of Plan and Finance - Ethiopia. Tigray Regional State: 5 Years Growth & Transformation Plan (2010/11 - 2014/15). Tech. Rep. Available at http://www.tigraionline.com/growth_transformation_plan_in_tigrai_state.pdf.
The National Regional Government of Oromia. Oromia Regional Accounts Part IV. Available at https://oromiabofed.org/index.php?option=com_content&view=category&layout=blog&id=45&Itemid=62.
Regional State Government Amhara. Amhara region GDP 1998–2003. Available at https://www.amharareg.gov.et.
Tilastokeskus - Statistics Finland’s free-of-charge statistical databases. Income and production by area, annually by Area, Transaction, Industry, Sector, Year and Information. Available at https://pxdata.stat.fi:443/PxWebPxWeb/pxweb/en/StatFin/StatFin__altp/statfin_altp_pxt_12bd.px/.
French National Institute of Statistics and Economic Studies. Produits intérieurs bruts régionaux et valeurs ajoutées régionales de 1990 à 2021 | Insee. Available at https://www.insee.fr/fr/statistiques/5020211.
National Statistics Office of Georgia (Geostat). Statistical Information by Regions of Georgia. Available at http://www.geostat.ge/regions/.
Statistische Ämter des Bundes und der Länder. Volkswirtschaftliche Gesamtrechnungen der Länder - Ausgabe 2010. Tech. Rep. (2010). Available at https://statistik.thueringen.de/webshop/pdf/2010/61002_2010_01.pdf.
Hellenic Statistical Authority - ELSTAT (Greece). Regional Accounts - Gross Value Added by Industry. Available at https://www.statistics.gr/en/statistics/-/publication/SEL45/-.
Fundacion para el Desarrollo de Guatemala. Indice de Competitividad Local. Available at https://www.fundesa.org.gt/indices-y-evaluaciones-de-pais/indice-de-competitividad-local.
Ministry of Statistics and Program Implementation | Government Of India. GSVA (gross state value added)/NSVA by economic activities. Available at https://mospi.gov.in/GSVA-NSVA.
Statistics Indonesia. Gross Regional Domestic Products (Industrial Orgins). Available at https://www.bps.go.id/subject/52/produk-domestik-regional-bruto-lapangan-usaha-.html#subjekViewTab3.
Statistical Centre of Iran. Provincial Gross Value Added. Available at https://www.amar.org.ir/.
Central Statistics Office Ireland. RAA06 - Gross Value Added. Available at https://data.cso.ie/table/RAA06.
iStat - Italian National Institute of Statistics. National Accounts: Regional main aggregates: Value added by industry. Available at http://dati.istat.it/Index.aspx?QueryId=11479&lang=en.
Cabinet Office - Government of Japan. Statistical Table: Prefectual Accounts. Available at https://www.esri.cao.go.jp/jp/sna/data/data_list/kenmin/files/files_kenmin.html.
Agency for Strategic planning and reforms of the Republic of Kazakhstan - Bureau of National statistics. Statistics of national accounts. Available at https://stat.gov.kz/en/industries/economy/national-accounts/dynamic-tables/.
Smailov, A. National Accounts of the Republic of Kazakhstan 2006–2010 statistical compendium. Tech. Rep., Agency on Statistics of the Republic of Kazakhstan, Astana (2012).
Smailov, A. National Accounts of the Repulic of Kazakhstan 2007 - 2011 statistical compendium. Tech. Rep., Agency on Statistics of the Republic of Kazakhstan, Astana (2013).
Kenya National Bureau of Statistics. Statistical Abtract (multiple years). Available at https://www.knbs.or.ke/publications/.
Kenya National Bureau of Statistics. Launch of the Gross County Product 2019 Report. Available at https://www.knbs.or.ke/launch-of-the-gross-county-product-2019-report/ (2019).
National Statistical Committee of the Kyrgyz Republic. Gross regional product (GRP) by types of economical activity in current prices. Available at http://www.stat.kg/en/opendata/category/4025/.
Central Statistical Bureau of Latvia. Value added and its share, total employment and employees by kind of economic activity and region and city (at current prices) 2000–2020. Available at https://data.stat.gov.lv:443/pxweb/en/OSP_PUB/START__VEK__IK__IKR/IKR020/.
Department of Statistics Malaysia. State Socioeconomic Report (various releases). Available at https://www.dosm.gov.my/portal-main/release-content/state-socioeconomic-report-2021.
National Statistics Office of Mongolia. Gross Domestic Product, by region, aimags and the Capital, by divisions (economic). Available at https://www.1212.mn/en/statistic/statcate/573052/table-view/DT_NSO_0500_021V2.
Haut-Commissariat au Plan du Maroc. Les comptes régionaux. Produit intérieur brut et dépenses de consommation finale des ménages 2020. Available at https://www.hcp.ma/Les-comptes-regionaux-Produit-interieur-brut-et-depenses-de-consommation-finale-des-menages-2020_a3572.html.
National Institute of Statistics Mozambique. Annuals (national accounts) – Provincial GDP. Available at http://www.ine.gov.mz/estatisticas/estatisticas-economicas/contas-nacionais/anuais-1.
Bikas Udhyami. Table 1: Provincial Annual Gross Value Added by Industrial Division. Available at https://nepalindata.com/media/resources/items/12/bPGDP_2077_78.xlsx.
Statistics Netherlands. StatLine - Production process, economic activity, regio; national accounts. Available at https://opendata.cbs.nl/statline/#/CBS/en/dataset/84419ENG/table?ts=1684321844860.
New Zealand Government. Regional gross domestic product: Year ended March 2020. Available at https://www.stats.govt.nz/information-releases/regional-gross-domestic-product-year-ended-march-2020.
National Bureau of Statistics - Nigeria. States Nominal Gross Domestic Product (2013–2017) - PHASE II. Available at https://nigerianstat.gov.ng/elibrary/read/947.
Republic of North Macedonia State Statistical Office. Publication: Regions of Republic of Macedonia. Available at https://www.stat.gov.mk/PublikaciiPoOblast_en.aspx?id=32&rbrObl=37.
Statistics Norway. Regional accounts, by industry (C) 2008 - 2020. Available at https://www.ssb.no/en/system/.
Statistics Norway. Digitized publications published 1977-96 - by subject: National Accounts by County (1973, 1976, 1980, 1983, 1986, 1992). Available at https://www.ssb.no/a/en/histstat/publications/ereg77-96_en.html#09.
Pakistan Bureau of Statistics. Household Integrated Economic Survey 2018/19 (HIES) Table 11: Percentage Distribution of Monthly Household Income by Source and Quintiles. Available at https://www.pbs.gov.pk/publication/household-integrated-economic-survey-hies-2018-19.
Pakistan Bureau of Statistics. Household Integrated Economic Survey - Table 11 (various prior editions). Available at https://www.pbs.gov.pk/search/node?keys=Household+Integrated+Economic+Survey+%28HIES%29.
Instituto Nacional de Estadística y Censo - Panama. Estimated Figures of Provincial Gross Domestic Product. Available at https://www.inec.gob.pa/publicaciones/Default3.aspx?ID_PUBLICACION=1032&ID_CATEGORIA=4&ID_SUBCATEGORIA=26.
Instituto Nacional de Estadística (INE) - Paraguay. Principales indicadores de ingresos por año de la encuesta, según departamento, 1997/98 - 2021. Serie comparable. Available at https://www.ine.gov.py/default.php?publicacion=5.
Instituto Nacional de Estadística e Informática (Peru). Gross Domestic Product by Department. Available at https://www.inei.gob.pe/estadisticas/indice-tematico/producto-bruto-interno-por-departamentos-9089/.
Instituto Nacional de Estadística e Informática (Peru). Cuentas Nacionales del Perú: Producto Bruto Interno por Departamentos 2001–2007. Available at https://www.inei.gob.pe/media/MenuRecursivo/publicaciones_digitales/Est/Lib0784/index.htm.
Philippine Statistics Authority. Gross Regional Domestic Product (GRDP). Available at https://psa.gov.ph/grdp.
Statistics Poland. Regional Accounts - Gross value added by PKD 2007 (NACE Rev. 2) sections groups. Available at https://bdl.stat.gov.pl/bdl/dane/podgrup/wymiary.
Statistics Portugal. Gross value added by NUTS 3 and by industry A3 (current prices; annual). https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_pesquisa&frm_accao=PESQUISAR&frm_show_page_num=1&frm_modo_pesquisa=PESQUISA_SIMPLES&frm_texto=Valor+acrescentado+bruto+por+NUTS+III+e+ramo+de+atividade+A3&frm_modo_texto=MODO_TEXTO_ALL&frm_data_ini=&frm_data_fim=&frm_tema=QUALQUER_TEMA&frm_area=o_ine_area_ContasNacionais.
National Institute of Statistics Romania. 5. Regional Gross Domestic Product. Available at http://statistici.insse.ro/shop/index.jsp?page=tempo2&lang=en&context=35.
Russian Federal State Statistics Service. National Accounts - Gross Regional Product. Available at https://eng.rosstat.gov.ru/folder/13913.
Statistical Office of the Slovak Republic. Gross value added. Available at http://statdat.statistics.sk/cognosext/cgi-bin/cognos.cgi?b_action=cognosViewer&ui.action=run&ui.object=storeID(%22i1357A4BF25EF40E1943195E2A720C621%22)&ui.name=Gross%20value%20added%20%5bnu3007rr%5d&run.outputFormat=&run.prompt=true&cv.header=false&ui.backURL=%2fcognosext%2fcps4%2fportlets%2fcommon%2fclose.html&run.outputLocale=en.
Statistical Office of the Republic of Slovenia. Gross value added at basic prices by activities, statistical regions, Slovenia, annually. Available at https://pxweb.stat.si:443/SiStatDataSiStatData/pxweb/en/Data/Data/0309254S.px/.
Statistics South Africa. Provincial GDP Figures. Available at https://www.statssa.gov.za/?page_id=735.
Statistics South Africa. Gross Domestic Product - Regional Estimates 2004 - 2013 (Tables 17-25). Tech. Rep. Available at http://www.statssa.gov.za/publications/P0441/P04413rdQuarter2014.pdf (2014).
Statistics South Africa. Gross Domestic Product - Annual estimates per region 1995-2004 (Tables 17–25). Tech. Rep. Available at http://www.statssa.gov.za/publications/P0441/P04413rdQuarter2005.pdf (2005).
Statistics Korea. Gross regional domestic product by economic activity. Available at https://kosis.kr/statHtml/statHtml.do?orgId=101&tblId=DT_1C81language=enconn_path=I3.
Spanish Statistical Office. Spanish Regional Accounts - 1980-96 Series & 1995-2004 Series. Available at https://www.ine.es/dyngs/INEbase/en/operacion.htm?c=Estadistica_C&cid=1254736167628menu=enlacesidp=1254735576581.
Central Bank of Sri Lanka. Economic and Social Statistics of Sri Lanka - Table 1.11 Provincial Gross Domestic Product by Industrial Origin at Current Market Prices. Available at https://www.cbsl.gov.lk/en/publications/other-publications/statistical-publications/economic-and-social-statistics-of-sri-lanka.
Statistics Sweden. Regional Accounts. Available at https://www.scb.se/en/finding-statistics/statistics-by-subject-area/national-accounts/national-accounts/regional-accounts/.
Federal Statistical Office - Switzerland. Gross domestic product per canton and region. Available at https://www.bfs.admin.ch/bfs/en/home/statistiken/volkswirtschaft/volkswirtschaftliche-gesamtrechnung/bruttoinlandprodukt-kanton.html.
Tanzania National Bureau of Statistics. Gross Domestic Product (GDP) - Annual National Accounts Publications. Available at https://www.nbs.go.tz/index.php/en/gross-domestic-product-gdp.
Thailand Office of the National Economic and Social Development Council. Gross Regional and Provincial Product (GPP). Available at https://www.nesdc.go.th/nesdb_en/more_news.php?cid=156&filename=index.
Turkish Statistical Institute. Regional Accounts. Available at https://biruni.tuik.gov.tr/medas/?kn=116&locale=en.
Office for National Statistics (UK). Regional gross value added (income approach). Available at https://www.ons.gov.uk/economy/grossvalueaddedgva/datasets/regionalgrossvalueaddedincomeapproach.
Office for National Statistics (UK). Historical Regional GDP 1968 to 1970 and 1971 to 1996. Available at https://www.ons.gov.uk/economy/grossvalueaddedgva/adhocs/005458historicalregionalgdp1968to1970and1971to1996.
State Statistics Service of Ukraine. Regional Statistics. Available at https://ukrstat.gov.ua/.
Uruguay Territory Observatory - OPP. Regional Gross Domestic Product by branches of activity. In millions of current pesos. Available at https://otu.opp.gub.uy/?q=listados/listados_datos_formato&id=2663&cant=0fecha=2014-01-01.
Bureau of Economic Analysis (USA). Regional Economic Accounts - Previously Published Estimates. Available at https://apps.bea.gov/regional/histdata/.
Statistics Agency under the President of the Republic of Uzbekistan. National Accounts. Available at https://stat.uz/en/official-statistics/national-accounts.
General Statistics Office of Vietnam. Socio-economic statistical data of 63 provinces and cities. Available at https://www.gso.gov.vn/en/data-and-statistics/2020/05/socio-economic-statistical-data-of-63-provinces-and-cities/ (2020).
General Statistics Office of Vietnam. GRDP at current prices by economic sector by Year, Indicators, and Economic sector (Unit: Bil. VND). Available at https://www.gso.gov.vn/ (2016).
Neri, F. The economic performance of the states and territories of australia: 1861–1992. Economic Record 74, 105–120, https://doi.org/10.1111/j.1475-4932.1998.tb01909.x (1998).
Vernon, J. D. Empalme series de pib regionales 1960–2001, base 1996. Technical report, Gobierno de CHile: Ministerio de Planificacion y Cooperacion (2004).
Lira, I. S. Disparidades, competitividad territorial y desarrollo local y regional en america latina. Technical report, Instituto Latinoamericano y del Caribe de Planificacion Economica Social (ILPES) (2003).
József, N. N. Fordulatra várva–a regionális egyenlötlenségek hullámai. Dövényi Zoltán–Schweitzer Ferenc: A földrajz dimenziói. MTA Földrajztudományi Kutatóintézet, Bp 141–158 (2005).
Resosudarmo, B., Y.et al. The East Asian experience: Indonesia. In The Dynamics of Regional Development: The Philippines in East Asia (Edward Elgar Publishing, 2007).
Felice, E. & Vecchi, G. Italy’s growth and decline, 1861–2011. Journal of Interdisciplinary History 45, 507–548, https://doi.org/10.1162/JINH_a_00757 (2015).
Nolintha, V. Cities, sezs and connectivity in major provinces of Laos. In Intra-and Inter-City Connectivity in the Mekong Region, BRC Research Report (Bangkok Research Center, 2011).
JICA. Data collection survey on economic development of the southern region in Lao People’s Democratic Republic. Technical report, International Development Center of Japan, Inc., Kaihatsu Management Consulting, Inc., IC Net Limited, Oriental Consultants Co., Ltd. (2012).
Cheng, M. H. The impact of ethnicity on the regional economic development of Malaysia. Master’s thesis, Graduate School of Social Sciences, Hiroshima University. (2011).
German-Soto, V. Generacion del producto interno bruto Mexicano por entidad federativa, 1940-1992. El Trimestre Economico 72, 617–653, https://doi.org/10.20430/ete.v72i287.574 (2005).
Arby, M. F. Some issues in the national income accounts of Pakistan (rebasing, quarterly and provincial accounts and growth accounting). Ph.D. thesis, Pakistan Institute of Development Economics (2008).
Manasan, R. G. & Mercado, R. G. Regional Economic Growth and Convergence in the Philippines: 1975 - 1999. Tech. Rep. DISCUSSION PAPER SERIES NO. 99-13, Philippine Institute for Development Studies (1999).
Ponomarenko, A. Chapter 10: Gross regional product for Russian regions: Compilation methods and preliminary results. In Regions: A Prism to View the Slavic-Eurasian World: Towards a Discipline of “Regionology”, 262–277 (Slavic Research Center, Hokkaido University, 2000).
Enflo, K., Henning, M. & Schön, L. Swedish regional gdp 1855–2000: Estimations and general trends in the Swedish regional system. In Research in Economic History, 47–89, https://doi.org/10.1108/S0363-3268(2014)0000030000 (Emerald Group Publishing Limited, 2014).
Karaca, O. Turkiye’de bolgeler arasi gelir farkliliklari: Yakinsama var mi. Turkiye Ekonomi Kurumu Tartisma Metni 7 (2004).
Mykhnenko, V. & Swain, A. Facing the global economic crisis: Ukraine’s uneven development and regional trajectories, 1990–2009. Technical Report, Economic geography research group symposia (2009).
García, M., Martínez-Galarraga, J. & Willebald, H. Crecimiento y estructura productiva regional en uruguay en la primera mitad del siglo xx: primeras aproximaciones y algunas hipótesis. Tech. Rep., instituto de economía, Universidad de la República (2014).
Nassef, A. F., Osman, O. M., Habib, S. K. & Ryhan, H. M. Human development report on education. UNDP (United Nations Development Programme). Available at http://www.undp.org.eg/ (1998).
Osman, O. M. & Ali, S. A. National human development report 2000/2001 Egypt. UNDP (United Nations Development Programme). Available at http://www.undp.org.eg/ (2000).
Vigilante, A. & Allam, S. T. Egypt human development report. UNDP (United Nations Development Programme). Available at http://www.undp.org.eg/ (2003).
Handoussa, H., Abdel Hay, M., Osman, M., Kandil, R. & Tobala, Z. Egypt human development report: Choosing decentralization for good governance. UNDP (United Nations Development Programme). Available at http://www.undp.org.eg/ (2004).
United Nations Human Development Program. Egypt human development report 2005: Choosing our future: towards a new social contract. Available at http://digitallibrary.un.org/record/602618 (2005).
United Nations Human Development Program. Egypt human development report 2008: Egypt’s social contract: the role of civil society. Available at http://digitallibrary.un.org/record/639484 (2008).
Ter-Nikoghosyan, V. et al. 2010 Egypt human development report: Youth in egypt. UNDP (United Nations Development Programme). Available at http://www.arab-hdr.org/ (2010).
Pleitex, W., Avalos de Trigueros, C., Valdez, M., Acevedo, C. & Trigueros, A. Evaluation of human development advances and challenges after a decade of changes. UNDP (United Nations Development Programme). Available at http://www.pnud.org.sv/ (2001).
Informe sobre desarrollo humano El Salvador, 2003: desafíos y opciones en tiempos de globalización. Available at http://digitallibrary.un.org/record/639570 (2003).
Informe sobre desarrollo humano El Salvador, 2005: una mirada al nuevo nosotros; el impacto de las migraciones. Available at http://digitallibrary.un.org/record/601854 (2005).
Hainsworth, G. & Hoang Dai Trang, N. El empleo en uno de los pueblos más trabajadores del mundo. UNDP (United Nations Development Programme). Available at http://www.pnud.org.sv/ (2008).
Pavlik, Z. El salvador 2010: Lineaminentos para un nuevo modelo de desarrollo. UNDP (United Nations Development Programme). Available at http://www.pnud.org.sv/ (2010).
United Nations Development Program. Informe sobre desarrollo humano El Salvador 2013. imaginar un nuevo país. hacerlo posible. diagnóstico y propuesta. (2013).
United Nations Development Programme. Honduras human development report 1999 on the pillars of human security and participation. Available at https://hdr.undp.org/content/report-pillars-human-security-and-participation (1999).
United Nations Development Programme. Informe sobre desarrollo humano Honduras 2002 (2002).
Calix, A., Castillo Perez, C., Sierra Fonseca, R., Leyva, H. & Murillo Selva, R. Culture human development. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/culture-human-development (2003).
Gallardo, G., Calderón, F., Morales, J. R., Sauma, P. & Dammert, L. Informe nacional sobre desarrollo humano 2006. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/informe-nacional-sobre-desarrollo-humano-2006 (2006).
United Nations Development Programme. Honduras human development report 2008/9: From social exclusion to a youth citizenship. Available at https://hdr.undp.org/content/social-exclusion-youth-citizenship (2009).
Mindaoudou, A., Oumarou, E. M. S., Hamza, H., Souleymane, A. & Boubacar, S. Informe sobre desarrollo humano Honduras 2011: Reducir la inequidad. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/reducir-la-inequidad (2011).
Saidova, G. et al. General human development report: Uzbekistan 1995. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/general-human-development-report-uzbekistan-1995 (1995).
Saidova, G., Olimdjonov, O., Bahramov, A., Isroilov, S. & Pogrebnyak, A. Role of the state in a period of transition to a market economy. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/role-state-period-transition-market-economy (1998).
United Nations Development Programme. General human development report Uzbekistan 1999. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/general-human-development-report-uzbekistan-1999 (1999).
Saidova, G. et al. Human development and SME. UNDP (United Nations Development Programme). Available at https://hdr.undp.org/content/human-development-sme (2000).
Saldova, G., Abduganleva, Y., Abdullaev, A., Akhmedov, T. & Anoshkina, V. Human development and decentralization. Available at https://hdr.undp.org/content/human-development-and-decentralization (2005).
United Nations Development Programme. Uzbekistan human development report 2006. Available at https://hdr.undp.org/content/uzbekistan-human-development-report-2006 (2006).
United Nations. International standard industrial classification of all economic activities (ISIC). Statistical Papers Series M No.4, Rev.4. Available at https://unstats.un.org/unsd/publication/seriesm/seriesm_4rev4e.pdf (2008).
Acknowledgements
We are grateful to Athene Cook and Amanda Lee for their excellent work compiling a first version of this database, and to Jan Borner, Irfan Mujahid, Yessengali Oskenbayev and Thang Dao Nguye for finding and providing specific sub-national data sets.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Author information
Authors and Affiliations
Contributions
R.C. and N.K. collected, harmonised and processed the raw data, building on an earlier version of DOSE. R.C. and N.K. created the custom shapefiles and helped presenting the results. M.Ko. supervised the collection of data, and assisted the data set harmonization and technical validation. M.Ko. and L.W. contributed to the date set’s structure and presentation. M.Ka. and L.W. proposed the generation of DOSE and advised the calculation of the different economic variables. R.C., N.K., M.Ko. and L.W. wrote the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Wenz, L., Carr, R.D., Kögel, N. et al. DOSE – Global data set of reported sub-national economic output. Sci Data 10, 425 (2023). https://doi.org/10.1038/s41597-023-02323-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41597-023-02323-8
- Springer Nature Limited
This article is cited by
-
Global economic impact of weather variability on the rich and the poor
Nature Sustainability (2024)
-
Climate damage projections beyond annual temperature
Nature Climate Change (2024)
-
Harmonized disposable income dataset for Europe at subnational level
Scientific Data (2024)
-
The economic commitment of climate change
Nature (2024)