Note: Charts current as of date above; text updated on January 6, 2021.
Incomplete information on the split between full and partial vaccination complicates the assessment of global vaccination progress. Most countries publish daily data on how many vaccine doses were administered and whether these doses initiated or completed a vaccination cycle (or both as in the case of single-dose protocols). Several countries, however, do not regularly publish such data.
This includes China, which given the sheer scale of its vaccination program has a large influence on the global number or, for that matter, any other aggregation of countries that it belongs to. The gaps between releases lead to an accumulation of “undocumented” doses that can become very large (undocumented in the sense that contemporaneous data on full versus partial vaccination is not available).
The difference between full and partial vaccination matters. Full vaccine efficacy is achieved some time after completing the final dose of the vaccine’s protocol – this could be the 1st dose (e.g. CanSino and Johnson & Johnson), the 3rd one (e.g. Abdala) or the 2nd one (most other vaccines). With more contagious, slightly more immune evasive and potentially somewhat more lethal variants circulating, full vaccination is more important than ever.
The charts below incorporate the very latest information from China and provide theoretical ranges for estimated rates to bridge any lack of recent information. To visualize divergences across countries, we cut the data by World Bank income group, World Bank region and UN subregion and show two types of information. First, we show what is currently know based on the last direct observations. These are current for most countries but dated for several others. Second, we show what we don’t know and calculate theoretical ranges of how undocumented doses could have affected the split under alternative strategies that either maximize full or partial vaccination. The actual values are likely to be somewhere within these ranges.
At the end of this post, find a more complete discussion about data challenges and the methods to deal with them. Let’s now proceed to the results.
Results by World Bank income classification
The state of global vaccination progress can be summed up as follows:
Results by World Bank region
Let’s also look at the results by region. We first use the World Bank’s classification of regions and then move to the UN geo-scheme.
Results by UN subregion
The UN geo-scheme divides the world into 22 subregions. This more disaggregated visualization drives home earlier points with greater granularity.
Annex: Digression on data issues
To assess global vaccination progress we need to tackle three data issues.
Data issue #1. No information on the split
Some countries do not publish ANY information on the split between those vaccinated fully or partially. This only applies to a small number of countries, which claim a tiny share in global vaccinations and will therefore not influence the results in a significant way. The note to each chart documents the number of countries, number of undocumented doses and the overall population these countries represent.
Data issue #2. Dated information on the split
Several countries publish dated information on the split. To correctly calculate vaccination progress, we need for a given date both the number of people fully vaccinated and those partially vaccinated (or alternatively those vaccinated with at least 1 dose). However, this information is not contemporaneously available for several countries. See the note to each chart, which shows the number of countries, the number of doses that are “undocumented” and the overall population they represent.
This data issues can significantly distort the global estimates depending on how dated the last information is and how much vaccination progress was made in the interim. As noted earlier, the main reason why this is a challenge is that China does not always publish contemporaneous information on the split, which given the scale and speed of its vaccination program influences the global estimates.
In light of the above, we calculate theoretical ranges for the current estimated values of the full and partial/at least 1 dose vaccination rates. These are calculated by allocating doses on the basis of strategies that maximize either partial vaccination (to expand reach) or full vaccination (to complete protocols and afford maximal protection). If needed, we censor the obtained value at the population maximum (and redistribute any excesses into either partial or full vaccination) and take the resulting value as the upper bound for the vaccination rate with at least 1 dose and the full vaccination rate, respectively.
The expanded estimates are illustrated in the charts by the contoured rectangles on the top of the bars. Where the latest data point is updated, the charts provide the actual latest information as well as the current estimate. In some cases, the expanded estimates are not shown since all vaccination data on the split is up-to-date.
Data issue #3. Internally inconsistent data on the split
The data reporting by several countries is not consistent with the mix of vaccines that they are using. The following relationship should hold at all times: doses_single_protocol = people_full_vaccinated + people_vaccinated (with at least 1 dose) – (total_vaccinations – total_boosters).
But for some countries the derived value for doses administered under 1-dose protocols (such as J&J and CanSino) is nonzero even though the country is not administering 1-dose vaccines. Conversely, for some countries, the data implies that such doses are zero even though 1-dose vaccines are being administered.
Our approach here is to go with the reported data. The inaccuracy is unlikely to influence the results by all that much as single-dose protocols represent a minor share in overall vaccinations.
Note: Thanks to Louis Kuijs and Tommy Wu of Oxford Economics for providing references to CanSino and the most recent China data.