In the case of East Asia and the Pacific, the data for China on hukou registration are not directly comparable with other sources of birth registration in the UNICEF global database. However, the available evidence seems to suggest that the level of birth registration (as measured by hukou registration) in China is at least as high as most other countries in the region. For this reason, there is practically no change in the regional figures of overall birth registration prevalence among children under age 5 and under age 1 if China were to be excluded due to lack of comparability of the data. However, since there are no comparable data for China on possession of a birth certificate (both for children under age 5 and under age 1), population coverage for this indicator for East Asia and the Pacific is below 50 per cent. Country-level assessment of progress For statistical purposes, complete birth registration coverage is defined as 99.5 per cent or higher. An assessment of progress was conducted for each country that had at least two comparable data sources, provided that the most recent data source was from 2010 or later. There were 156 countries for which an assessment of progress was possible. Of these 156 countries, 67 were considered to have already met the target of complete birth registration coverage. For each country, the average annual rate of change (AARC) was calculated using the two most recent data points. Given that evidence suggests that progress slows as levels of birth registration approach 100 per cent, a country’s observed AARC may no longer be plausible when its projected levels get higher. To adjust for this, the projected birth registration prevalence for each country followed the country’s trajectory, starting from the latest data point until the prevalence of birth registration reached a level at which its observed AARC has never been recorded (at least among the countries with available trend data). At this point, the projection scenario continues forward at the highest AARC observed by any country within that tier. The following tiers were established (based on starting levels of birth registration): 0-24 per cent, 25-74 per cent, 75-84 per cent, 85-94 per cent and above 95 per cent. On the basis of the projections, countries were assigned to one of the following two categories, which represent their progress towards the target of universal birth registration by 2030: • On track – the projected prevalence of birth registration after applying the method described above is expected to reach a level of 99.5 per cent or above by 2030. • Not on track – the projected prevalence of birth registration after applying the method described above is not expected to reach a level of 99.5 per cent or above by 2030. These analyses need to be interpreted with some degree of caution since the changes observed that were used to determine a country’s progress may be within confidence intervals. Global trend analysis For the global-level trend analysis, countries that had at least two comparable data sources were included if they met the following criteria: the earliest data point corresponds to a range between 1999 and 2005, a mid-point corresponds to a range between 2006 and 2010, and/or the most recent data point corresponds to a range between 2011 and 2018. An exception to this rule was made for 29 countries that had an available data point that was one or two years outside of these ranges. For China, trend data for the years 2000, 2010 and 2015 on children’s hukou registration were also available and therefore included in the calculation of the global AARC figures. For countries that had only two available data points to use for the trend analysis, the missing data point (either the earliest or a mid-point) was imputed by applying the country’s observed AARC between the two available data points, based on the assumption of a straight-line trajectory. This methodology was applied to roughly one third of the countries with trend data. This was done in order to maximize the available country trend data. Regional trend analysis For the regional trend analysis, the same set of countries included in the global analysis was utilized. Each region was assessed separately in terms of the availability of country trend data within the region. For example, if the majority of the available country trend data in a region corresponded to the years 2000, 2011 and 2016, then the regional AARC figures were calculated as a 11-year trend (to capture the change that occurred between 2000 and 2011) and a 5-year trend (to capture the more recent change that occurred between 2011 and 2016). In this sense, each region’s AARC figures have been tailored to reflect progress observed in that region over a specific period of time. As was done in the case of the global trend estimates, the value for ‘today’ (calculated on the basis of all countries with available data in the region between 2010 and 2018 and not just those with trend data) was used as the starting point. Technical annex 37