Coding rules

In order to ensure consistency for the data in WeSIS and in the documentation in WeSISpedia, several coding rules, documentation standards and conceptual agreements have been established.

Technical rules

Date time

Dates, e.g., the introduction of laws, election dates etc. must be entered as YYYY.MM.DD.

For date times the following rules apply in descending order:

1. Whenever possible, add the complete date.
2. If an event (like an election) took place on more than one day enter the first day of the period.
3. If the day is unknown, enter YYYY.MM. While importing data to WeSIS an internal flag will be set that the day is unknown.
4. If even the month is unknown, just enter YYYY. Internally, a flag will be set that both the day and month are unknown.
5. The default reference date for all yearly data is December 31, if not explicitly stated otherwise on the indicator pages.

Country codes

WeSIS makes use of the entity codes defined by the Corrrelates of War project (COW). See the entire list and additional notes on the country codes page.

Missing values

WeSIS has a generic approach of flagging "missing data" using an exclamation mark (cf. file upload guide), but there is a recommended set of numerical codes to express different kinds of "missingness".

Technical variable names

Within WeSIS every indicator has a technical name. The topics define the initial abbreviation followed by an underscore. Add a meaningful suffix afterwards using underscores as well, e.g., polnat_election_date_1st_chamber (not polnat.electiondate) for an indicator capturing the election date of the 1st chamber. The name must be unique and may encompass subtopics, units and sources if necessary.

A recommended pattern reads as follows: topic_subcategory1_subcategory2_name_unit_source (for example: health_hospitals_beds_per1000_ibge or econrel_trade_total_import_usd_wto).

Notes: Try to not exceed the limit of 128 characters for the full name including any underscore. Stick to lower-case letters for the technical name (already uploaded data will not be affected however) and avoid "special" characters (for example, +, ?, :, or /).

Social policies

Topic	Prefix
Old age and survivors	old_
Labour and labour market	labor_
Health and long-term care	health_
Education and training	edu_
Family policies	fam_
Gender aspects	gend_
Immigrant social rights	isr_

Domestic conditions

Topic	Prefix
Policy legacies	polleg_
Economic and financial factors	econnat_
Political factors	polnat_
Social structure	socstr_
Culture	cult_
Geography	geo_
Sustainable development goals	sdg_

Interdependencies and relations

Topic	Prefix
Communication	comm_
Political institutional linkages	polrel_
Economic relations	econrel_
Migration	migra_
Violent conflicts	confl_

Conceptual agreements

Introduction dates

In principle, three dates are important: the date of adoption by (political) actors, the date a law comes into force (de jure) and the actual implementation (de facto). Beyond OECD countries and in some policy fields (especially in health care) there is often a huge lag between coming into force and the actual implementation. Yet, the implementation depends on a large number of factors that are difficult to capture. Coding the adoption, in turn, may be interesting from a political science point of view (e.g., partisan theory).

For WeSIS it was agreed that

1. the date a law comes into force (de jure) shall be recorded.
2. Amendments that extend the application of a law to different status groups or policy fields shall be recorded as if they were new.
3. The date of adoption or information regarding de facto implementation may be recorded if the information is easy to obtain or already known ("nice to know"). The responsibility to collect such information lies with the projects, though.

Regional fragmentation

In some countries, social policies and welfare schemes are regulated at a certain subnational level, at times leading to substantial variation between sub-national entities, particularly regarding inclusiveness and scope of benefits. Hence, in comparative analyses the question arises which subnational unit is used to "represent" the focal country. Possible solutions include the most populous entity today, the most populous entity at each point in time, the most populous entity at the start of the time period, the most influential entity (e.g., the first entity that industrialized and was therefore the first to develop a certain law or policy scheme). Each option comes with advantages and disadvantages though regarding

• comparability, e.g., when using the most populous entity at each point in time, a jump in the time-series may be due to switching the entity rather than actual policy changes;
• representativeness, e.g., the most popoulous state does not reflect the "typical" inclusiveness or scope of the policy or scheme per se but rather represents an "outlier" in terms of regional variation;
• consistency, e.g., in case there are breaks due to changes in statehood like colonies becoming independent states or regime break down;
• resources, e.g., the decision how many subnational entities need to be coded to (better) grasp the regional variation.

For WeSIS it was agreed that

1. the default is to collect data for the most populous entity today.
2. Projects may deviate from this rule for reasons grounded in the specificity of the focal policy or scheme. If they do so, they are obliged to justify and document their decision.
3. If they deviate from the default, projects are still encouraged to invest resources and collect data for the most populous entity today which already gives first hints about the actual variation.

Any deviation from the default has to be documented in WeSISpedia as follows:

• The indicator description shall include a note that regional fragmentation is present.
• The section on "Coding rules" shall state which subnational entity is used to "represent" the focal country in the indicator's time-series.
• When preparing the data for upload to WeSIS, the optional column "comment" shall be used to state the presence of regional fragmentation, e.g., "There is regional fragmentation. Data refers to entity XYZ".

If the projects invest the resources to also code additional entities (e.g. the most popoulos entity today), they have to decide which data is used to “represent” the focal country in the main indicator’s time-series. They can then upload further indicators containing only the data for the additional entities and document and name them accordingly. For example,

• the main indicator may be assigned the technical variable name labor_dis_not_per and be documented as "Legally mandated notice period" including California as the "representative" entity for the US,
• while a second indicator may be assigned the technical variable name labor_dis_not_per_florida containing only the data on Florida, and be documented as "Legally mandated notice period (Florida)" by describing the rationale of this indicator including cross-references to the main indicator.

This way, researches may later decide to download both – the main and any secondary indicator – to "replace" the time-series of the main indicator with data from other entities if it better fits their research question.

Spelling

BE or AE

In WeSISpedia American English shall be used.

Capitalization

Indicator names (and hence page titles) shall not be capitalized. For general guidelines please refer to Wikipedia's Manual of Style as well.

Citations

WeSISpedia applies the citation style of the American Political Science Review in line with the American Political Science Association's official style guide. The style uses the author-date system for in-text citations. It is used in the following format: (Author Year, Pages). There is no comma between the author and the year whereas mass citations are separated with a semicolon.

The citation style is readily available as a template in Citavi and RefWorks and can be downloaded for Zotero, EndNote or Mendeley.

Below you will find some examples for referencing common types of publication. Please refer to the APSA Style Guide for further information:

• Edited Volumes: Castles, Francis G., Stephan Leibfried, Jane Lewis, Herbert Obinger, and Christopher Pierson, eds. 2010. The Oxford Handbook of the Welfare State. Oxford: Oxford University Press.
• Monograph: Esping-Andersen, Gøsta. 1990. The Three Worlds of Welfare Capitalism. Princeton: Princeton University Press.
• Report (Monograph): OECD. 2006. Taxing Wages 2004/2005. Special Feature: Part-Time Work and Taxing Wages. Paris: OECD Publishing.
• Journal Article: Schmitt, Carina. 2014. "The Diffusion of Privatization in Europe: Political Affinity or Economic Competition?" Public Administration 92 (3): 615-635.
• Working Paper: Schmitt, Carina, and Herbert Obinger. 2012. Policy Diffusion and Social Rights in Advanced Democracies 1960-2000. Bremen: Zentrum für Sozialpolitik. ZeS-Working Paper, 02/2012.