Long-term population estimates are usually based on decennial (or roughly decennial), census-type data, especially those that go beyond the aggregate or national level. The lack of annual population datasets, however, restricts the range of analysis and therefore our knowledge of populations’ behaviour over time. This paper contributes to this scant literature by providing two new long-term, annually and spatially disaggregated population estimates for Spain, a key case study. We use a set of official sources and propose a straightforward method to estimate Spanish annual populations at two levels of spatial disaggregation, the provinces (NUTS III) and their capital cities (a selection of LAU II/NUTS V), between 1900 and 2011. For the cities, our estimates account for changes in boundaries. Our proposed method is especially suitable for countries with a lack of long-term, systematic, historical migration statistics, a common issue. To test the robustness of our estimates, some comparisons are made with previous more aggregated-level or decennial census-based estimates, and a preliminary analysis of the evolution of disaggregated Spanish population is presented.

中文翻译：

---

按空间分类人口的年度估计：西班牙，1900-2011年

长期人口估计通常基于十年（或大约十年）的普查类型数据，尤其是那些超出总水平或国家水平的数据。然而，缺乏年度人口数据集限制了分析的范围，因此限制了我们对一段时间内人口行为的了解。本文通过提供两个重要的案例研究，为西班牙提供了两个新的长期，年度和空间分类的人口估计数，从而为这一稀缺的文献做出了贡献。我们使用了一组官方资源，并提出了一种简单的方法来估算空间分布的两个层次上的西班牙年度人口，即省份（NUTS III）和其首都城市（部分LAU II / NUTS V），在1900年至2011年之间。对于城市，我们的估算考虑了边界的变化。我们提出的方法特别适用于缺乏长期，系统，历史移民统计数据且普遍存在的国家。为了检验我们的估计的稳健性，我们将其与以前基于汇总水平或十年期人口普查的估计进行了一些比较，并对西班牙分类人口的演变进行了初步分析。