Demographic change occurs when individuals arrive from elsewhere and couples bear children, adding new members to the population, whereas others die or depart to other places, which depletes it. Because these demographic events affect the population of different sexes and ages, they determine both the number and structure of a population. The main objective of every demographic prospect is to determine how births, deaths and migrations will occur during the next year. The methodology adopted for this third round of INSTAT’s population projections – the cohort component model – calculates these events for each year and age-sex group under the assumption of plausible future trends in mortality, fertility and migration in the future. These events are subtracted and added to the population in the beginning of the year, providing the population by age and sex at the start of the next year, and so on.

The number of demographic events depends on both the population structure and demographic behaviour (i.e. the number of births depends on the number of women in childbearing age and their rate of childbearing). As population structure and demographic behaviours are known at base-line (from the Census 2011 and the previous analysis, respectively), future trends in demographic behaviour were projected based on hypotheses that rely on our knowledge of the demographic situation, past trends, the socio-economic and political context and international comparisons (see next section).

The population by age and sex at base-line of the projections, 1st january 2011, is estimated by INSTAT through retropolation of the enumerated population on 1st October 2011 (taking into account recent emigration and immigration, as well as the rate of under-enumeration). The national population by sex and single age is projected iteratively year after year in the following manner, starting with the population at 1st January:

• The number of deaths is estimated for the first six months in applying sex and age-specific death rates to the population on 1st January; the deaths are subtracted from population to obtain the survivors at mid-year (30 June);
• The annual number of emigrants by age and sex are estimated from this mid-year population using emigration rates, and these events are then subtracted; an assumed number of immigrants by sex is redistributed by age and added to the population.
• The number of deaths for the second semester is then estimated and subtracted from this revised mid-year population, yielding the number of survivors on 31st December; • The number of births is estimated in multiplying age-specific birth rates to the average number of females in reproductive age; infant deaths and emigrants until the end of the year are estimated and subtracted from the number of births;
• To project the population on 1st January of the following year, each cohort is moved forward by one year of age to account for population ageing, and the number of surviving births within the country constitutes the first age-group.

Prefectural population by sex and 5-year age groups are projected iteratively and simultaneously over 5-year intervals using a cohort component multiregional migration-pool model, which was adapted from the multiregional model proposed by the U.N. (united Nations 1992). The procedure involves the following steps for each prefecture:

• The sex and age specific population at the start of the projections interval are forward-survived by 5 years,
• Surviving emigrants (abroad) and out-migrants (to other prefectures) are successively estimated using migrant transition rates (net of mortality, and international migration in the case of out-migration) which are applied to the population at the end of the interval. Emigrants are subtracted from the population and out-migrants are pooled by age and sex, providing the number of in-migrants to redistribute.
• The pooled number of in-migrants, as well as an assumed number of immigrants (from the national level projections) are added in the end. These were redistributed by prefecture according to their respective geographic distribution (by age and sex) in 2006-2011.
• Births are estimated in applying age-specific fertility rates to the average population in the beginning and end of the 5-year interval and are forward-survived to the end of the interval
• To obtain the population on 1st January at the start of the following 5-year interval, each cohort is moved forward by five years of age to account for population ageing; the number of surviving births within Albania constitutes the first age group.
• The sex- and age-specific populations and demographic events at the prefectural level are constrained to sum up to the totals projected by the national projections.