Course:Cons452/PEN

Poverty and Environment Network (PEN) by CIFOR

Description

The goal of the PEN network is to provide uniform socio-economic and environmental data at household and village levels in rural areas of developing countries. More specifically, it was designed to investigate the relative contributions of agricultural, forest and non-forest ("environmental") income. It contains data collected on the quantities of forest products (including foods) that are used by households. Created by the Centre for International Forestry Research (CIFOR), a key purpose of this dataset is to provide high-quality data on forest use and forest-poverty interactions. PEN is responding to a lack of research around the role that forests play in poverty alleviation in the tropics. The dataset comprises 8,301 households in 334 villages located in 24 countries in Asia, Africa and Latin America. The data were collected by 33 "partners", mostly PhD students. The dataset includes 3 types of quantitative surveys:

1. Village Surveys
   • The first village survey (V1) - conducted at the beginning of the fieldwork to get background information on the villages
   • The second village survey (V2) - conducted the end of the fieldwork to get information for the 12-month period covered by the surveys
2. Annual Household Surveys
   • First annual household survey (A1) - collected basic household information including demographics, assets, forest-related information and was carried out at the beginning of the survey period
   • Second annual household survey (A2) - collected information for the 12-month period covered by the surveys and was done at the end of the survey period
3. Quarterly Household Surveys
   • For income information (Q1, Q2, Q3, Q4)

The detailed household income surveys are a key component of the PEN dataset. They were done "quarterly" (that is, 4 times in one year) for two main reasons: asking more frequently improves the accuracy of recall, and it also allows seasonal variation to be captured. This matters a lot in the case of forest use as it can help us see how much forests may be used as seasonal "gap-fillers". In other words, in times of the year when other harvests are particularly scarce, they want to know if people are going to the forest more to acquire resources. Keep this seasonality aspect in mind when looking at the data.

What type of data is available: products acquired by the household (an incredibly detailed list of products is available, including: wild-harvested products from the forest, bush meat, processed goods, non-forest products, etc...), crops harvested, minerals, land tenure arrangements, income, type of work, mode of transport, household assets, agricultural inputs, livestock, and more.

Site-Selection: To have been included in this dataset, sites must have at least a minimum level of forest-dependence. PEN also requires all of its partners to choose villages and households that are representative, in other words that are not special cases or outliers with really unique circumstances. Another requirement is that the sites capture some variability, especially in the following areas: market access/remoteness, forest abundance, income levels, and institutional aspects. A minimum of 100 households in a village was also required. There is a PDF available for download in this dataset that outlines the codes of the different countries where data was collected. This code is the beginning of the SiteID codes, and so in that way you can isolate which data came from which country. However, more detailed geo-locating is not possible with this dataset.

Metadata

Metadata Component	Description
Theme	Forests and human well-being
Source	Centre for International Forestry Research (CIFOR)
Purpose	To collect detailed, high-quality data on forest use in a variety of tropical and subtropical settings
Time Frame	2004-2010
File Type	socioeconomic (survey)
File Format	.csv
Structure	One household per row, one village per row
Projection and coordinate system	N/A
Extent	Asia, Africa and Latin America
Resolution or scale	household, degree

Common Problems

• There are three partners (10101, 10203, and 10301) who, because of various particular circumstances, only conducted three of the four income surveys. In addition, 598 of the households missed out on one of the quarterly surveys. These are still in the database, but households missing more than one quarter were removed. The partners adjusted their recall periods to account for this which makes the surveys comparable with the rest, so the sum of the income collected in the three surveys in the household's annual income. Keep this in mind when thinking about annual income data.
• 2 partners had already used Version 3 of the PEN survey before they settled on Version 4. There were a few changes. For those items where they did not have the information "-8" has been put, which means "not applicable"
• Be careful to understand the exact meanings of the terms and variables. Please see the PEN Technical Guidelines PDF (downloadable where all the other data files are) for explicit definitions of what they mean when they use general terms like forest, agroforest, or land-tenure. Please see the PEN codes document (currently on version 7.7) for all variable codes.
• Codes for this dataset can be found in 3 different documents: the PEN codes (codebook); PEN country codes, partner codes, & names; and then also the PEN Data Introduction. Perhaps more surprisingly, the smaller document PEN Data Introduction (Careful: it is different than the PEN Technical Guidelines) contains the aggregate product codes, which are not found elsewhere).

Downloading Instructions

1. Go to: https://data.cifor.org/dataset.xhtml?persistentId=doi:10.17528/CIFOR/DATA.00021
2. There are a total of 69 data files available for download in a list at the bottom of the PEN dataset page. You can use the filters to help you select the data you are looking for. 11 of the files are document files, which are PDFs that contain information on variable labels, codes used, questionnaire used, technical guidelines, and site locations. You will want to download the PDF(s) that correspond to the variable names of the data file you want to use. One of the files is a map of the sites where they collected data. 57 files are tabular data, and under each file heading there is a concise description outlining which variables are contained within that file. If you would like to, you may select the checkboxes on the far left hand side to select more than one dataset at a time to download. The download button for the multi-file download is at the top right hand side of the box. The naming convention of the 57 data files is: module_section_content, and each section of the questionnaire is its own file.
3. Once you have selected your desired file(s), click the download button. A dropdown will prompt you to select whichever file type you would like.
4. A pop-up will appear asking you to enter your information and reason for the download. You do not need to make an account, but you do need to fill in this information if you want the download. The download occurs immediately once you have filled in the form.
5. If it is more than one file it will be in a zipped folder, otherwise it will be simply the file you requested (.pdf, .csv, etc...).

Restrictions on Use

These data and documents are licensed under a Creative Commons Attribution 4.0 International license. You may copy, distribute and transmit the data as long as you acknowledge the source through proper data citation. By clicking on the title hyperlink of each data file, you will be taken to a page where you are given both the dataset citation as well as the file citation already made.