I can’t find a volume.

We keep an up-to-date list of all of the digitized volumes here. If there is a “1” in the Missing column that means we are either in the process of scanning it or have not scanned it yet.

A page is missing/blurry.

Please use the contact us form to let us know. We will add it to the list of known issues to address.

You should add ___ to your collection.

Please use the contact form to suggest additional series, types of statistics, that would be helpful for you. We will do our best to continuously add/update available resources.

I contributed to the project. I want my name added to the RA list.

Please use the contact form to let us know.  Our sincerest apologies for the oversight.

How reliable are these statistics? Weren’t most of them recorded by colonial or authoritarian governments? Why should we trust them?

Some statistics are likely to be more reliable than others. In ongoing research, we discuss this question in greater detail, but statistics that were regularly produced, that were widely used by key ministries, and were subject to independent confirmation are likely to be the most reliable. For most years, budgetary data are illustrative of these characteristics. Statistics that were irregularly produced, associated with less important ministries, fundamentally biased in their construction, and that were difficult for an independent source to confirm are probably less reliable. For example, because of huge disparities in reporting between colonial citizens and subjects, in most countries in the MENAHDA dataset estimates of mortality statistics, particularly for colonial subjects, are likely to be significantly underestimated.

I am working with these statistics for the first time, can you please provide some suggestions?

Work from the most recent edition backwards in time.  More recent additions are more likely to report authoritative statistics in the case of an error or an estimation in an earlier edition.

Pay careful attention to units! Sometimes reported units change even within the same nominal statistic.

Statistical series tend to get much more aggregated around World War II and to stay quite aggregated until the early 1980s when often they start to become more detailed again. You may need to consult archival sources to find more disaggregated statistics.

Wars and other major political crises often disrupt the reporting and quality of official statistics. Some gaps in the publications, or aggregation in the reported statistics, coincide with the onset of a political crisis. Contextual knowledge is essential for evaluating gaps, changes in the reported units or quality of even the same series of statistical tables.

Where can I find historic maps from the region?
Why didn’t you OCR these volumes?

We made this determination based on logistical, technical, and practical concerns. Logistically it is difficult to know which volumes or sections of volumes would be most useful to researchers and digitizing each page of every volume would have been incredibly time-consuming and costly. A second challenge concerned the format of the tables which often involve Arabic and at least one other European language (usually French or English). In our experience, existing OCR technologies struggled to extract content given the complicated formatting and multiple languages regularly employed across the volumes in the dataverse. A final concern related to the reliability of the statistics reported in the volumes. As we note in ongoing work, there is significant variation in the reliability of the statistics reported across countries and years for each country in the Dataverse. Contextual knowledge is essential for making a discerning judgement about which statistics are reliable. Our judgement was that this was a decision that researchers needed to make for themselves.

What geographic units are included in these volumes?

Generally speaking all volumes included tables that report sub-national variation for key outcomes: population, schooling, roads, infrastructure, etc. However because the volumes were already large, when they are reported sub-national units tend to be large typically the Admin1 level: i.e governorates, or provinces. Typically earlier volumes include more fine grained units although this level of detail is usually not reported after 1945.