Data and Codes for: The Origins of Italian Human Capital Divides: New evidence from Marriage Signatures, ca. 1815
Principal Investigator(s): View help for Principal Investigator(s) Marco Martinez, Sant’Anna School of Advanced Studies
Version: View help for Version V2
Name | File Type | Size | Last Modified |
---|---|---|---|
CODES | 05/14/2023 03:44:AM | ||
DATA | 03/02/2023 06:00:AM | ||
|
text/csv | 4.7 MB | 05/13/2023 11:33:PM |
|
application/rtf | 4.7 KB | 05/16/2021 12:29:PM |
|
application/pdf | 11.3 MB | 05/13/2023 11:37:PM |
Project Citation:
Martinez, Marco. Data and Codes for: The Origins of Italian Human Capital Divides: New evidence from Marriage Signatures, ca. 1815. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2024-06-17. https://doi.org/10.3886/E121862V2
Project Description
Summary:
View help for Summary
This paper quantitatively assesses to what extent signatures in marriage certificates can inform about literacy rates in the past. The new figures are based on a novel and balanced random sample of marriage certificates for Italy in 1815. Such estimates are compared to all alternative sources available for close years. The interdisciplinary empirical methodology is as important as the results. Two main findings emerge. First, marriage certificates, although recorded from different pre-unitary states, can accurately measure literacy rates only when accompanied by rigorous sampling procedures. Indeed, the proposed empirical approach allows to go from local to aggregate estimates that are in line with other estimates for the period. Second, North vs Centre-South divides in 1815 are lower than previously suggested.
Scope of Project
Subject Terms:
View help for Subject Terms
regional divides;
education;
pre-industrial Italy;
pre-unitary Italy;
Napoleonic reforms
Geographic Coverage:
View help for Geographic Coverage
Italy
Time Period(s):
View help for Time Period(s)
1814 – 1816 (1815)
Data Type(s):
View help for Data Type(s)
administrative records data;
geographic information system (GIS) data
Methodology
Sampling:
View help for Sampling
I use a one-stage clustered sampling stratified by region scheme: I randomly select a number of villages within the village list of each archival unit. Then I weight the total sample size by the regional population weight and the number of contracts to have in each archive over the total of the region, and start collecting the required marriage certificates for each archival unit.
Data Source:
View help for Data Source
The pictures of the marriage certificates can be retrieved in:
http://dl.antenati.san.beniculturali.it/?q=gallery
Under the headings "Stato Civile Napoleonico" and "Stato Civile della Restaurazione".
http://dl.antenati.san.beniculturali.it/?q=gallery
Under the headings "Stato Civile Napoleonico" and "Stato Civile della Restaurazione".
Collection Mode(s):
View help for Collection Mode(s)
record abstracts;
web scraping
Scales:
View help for Scales
The literacy rates are based on a binary scale: either able to sign, or not able to sign. Sometimes, the enumerators explicitly mentioned that the respondents declared to be illiterate.
Weights:
View help for Weights
The sample should have a balanced picture of the Italian population, so that inferences are representative and not biased towards overrepresented areas. However, the marriage certificates digitised online are incomplete, and some areas of the North have fewer archives than their population suggests. Since by construction the sampling is stratified by archive, this would over-represent areas where the data collection is complete.
For this reason, I construct three weights: the first weight corrects for within-archive variability in the number of contracts, the second corrects for the between-region variability in archive coverage, and the third weight for the share of urban areas in the region. The 1815 population estimates were drawn from \cite{mariella2020reconstruction}, whose data reconstruction can be considered the most updated and complete, and is constructed using a wealth of micro population studies. I then aggregated at the level of unitary region in the Centre-North, and at the level of pre-unitary provinces in the South. I compared the resulting 1815 estimation with a similar estimation derived from other sources (ASI (1853), Castiglioni (1862)). The population in urban areas was derived from the 1800 estimations of Malanima (2016), an updated version of the database created in Malanima (1998). I consider 9'000 in 1800 as the minimum population threshold for larger towns in 1815, and 4'000 as minimum threshold for smaller towns. I also assume that when town size in 1800 is smaller than 5'000, it is 4'000. However, the share of smaller towns in the 1861 population census was much higher than the one derived by Malanima (2016), who assumes to have population 0 to towns that had less than 5'000 inhabitants in 1800. Assuming smaller towns to have a population of 4'000 partially corrected the discrepancy between the census and the sample urban-rural weights.
For this reason, I construct three weights: the first weight corrects for within-archive variability in the number of contracts, the second corrects for the between-region variability in archive coverage, and the third weight for the share of urban areas in the region. The 1815 population estimates were drawn from \cite{mariella2020reconstruction}, whose data reconstruction can be considered the most updated and complete, and is constructed using a wealth of micro population studies. I then aggregated at the level of unitary region in the Centre-North, and at the level of pre-unitary provinces in the South. I compared the resulting 1815 estimation with a similar estimation derived from other sources (ASI (1853), Castiglioni (1862)). The population in urban areas was derived from the 1800 estimations of Malanima (2016), an updated version of the database created in Malanima (1998). I consider 9'000 in 1800 as the minimum population threshold for larger towns in 1815, and 4'000 as minimum threshold for smaller towns. I also assume that when town size in 1800 is smaller than 5'000, it is 4'000. However, the share of smaller towns in the 1861 population census was much higher than the one derived by Malanima (2016), who assumes to have population 0 to towns that had less than 5'000 inhabitants in 1800. Assuming smaller towns to have a population of 4'000 partially corrected the discrepancy between the census and the sample urban-rural weights.
Unit(s) of Observation:
View help for Unit(s) of Observation
Each State Archive, or archival unit, has a list of ``villages'', or \emph{paesi}. These are often municipalities, but in some cases correspond to smaller villages or even parishes. Large cities (Palermo, Naples, and Turin) are considered separate archival units so that their larger population size does not overshadow the literacy rates of rural areas. Obtaining a list of the villages from the online platform is straightforward, but it would be very difficult to obtain an ordered list of marriage certificates within each archival unit, because in some cases there are more than one contract per scanned page, and in other cases there is a contract every two or even four pages. For this reason, villages are the unit of observation for sampling purposes, but the unit of analysis is the individual spouse (bride and groom).
Geographic Unit:
View help for Geographic Unit
Unitary region for the North and Centre of Italy; pre-unitary province for Southern Italy
Related Publications
Published Versions
Report a Problem
Found a serious problem with the data, such as disclosure risk or copyrighted content? Let us know.
This material is distributed exactly as it arrived from the data depositor. ICPSR has not checked or processed this material. Users should consult the investigator(s) if further information is desired.