Page 35 - Revista del Centro de Investigación y Desarrollo del INEI - Economía, Sociedad y Estadística N° 9

3. DATA AND METHODOLOGY Finally, the following will also be included as controls:

Data: - Rural population rate by the district in 1981.

- The logarithm of the population of the district in
The entire study will be conducted using secondary
data. In particular, the four population censuses of 1981.
Peru will be used: - Regional Fixed Effects.
- Dummy variables, if it is provincial capital, if it is
- VIII Population Census and Housing III 1981 .
1
department capital, and if it is Metropolitan Lima.
- IX Population Census and IV Housing 1993.
- XI Population Census and Housing VI 2007.
- XII Population Census and VII Housing 2017.

The indicators will be constructed:

- Immigration Rate, as the percentage of the
population, residing in the origin district but born
in another district divided among the people residing
in said district.
- Emigration Rate, as the percentage of the population 4. RESULTS:
that was born in the origin district and resides in
any other district divided among the population Typology:
that was born in origin district.
In this case, the information from the four censuses is
On the other hand, to study what factors are associated combined, and a typology will be constructed with the
with aging or rejuvenation in the district: combinations. In each census, three categories will be
generated: Sender, Receiver, and Neither Sender nor
- Sum of Older Adults (65 or older) in the district. Receiver (Ni-Ni).
- Adult Sum (40-64) in the district.
- Mostly Sender, if the district has a district emigration
- Sum of Youth (18-39) in the district. rate higher than the country’s average.
- Sum of Children and adolescents (under 18) in the - Mostly Receiver, if the district has a district
district. immigration rate higher than the country’s average.
- Neither Sender nor Receiver (Ni-Ni), if it is below in
The following variables are used as control variable or
determinants of migration patterns. both. That is, it is a district with low exit and entry
mobility.
- Altitude and average slope.
This classification was chosen using “Cluster Analysis”
- Dominant natural region (using the definition of using the combination of the Emigration Rate and
Pulgar-Vidal). Immigration Rate. The results of this classification for
the four censuses are presented in Table 1.
- Distance to the National Capital (Metropolitan Lima).
- Distance to the regional capital. Then, these three categories are combined for the four
censuses, which are summarized in the following 7,
- Distance to the provincial capital. shown in Table 2.

1 There is no information available of 3 regions: Apurímac, Loreto and San Martín, for these cases, the data were imputed using the
1993 information in combination with that of the other regions in 1981.
35