How to use?
This section contains information on how to use and interpret the information from the database.
DV - dependent (outcome)variable
IDV - independent (explanatory) variable
All hypotheses and findings are reported in terms of better compliance (shorter transposition time, shorter transposition delays, less infringement procedures in a year, etc.) and not in terms of the original dependent variables used in the studies.
How to search?
You can search the database by selecting options from the drop-down menus. For example, you can restrict the results from the search only to articles that focus on the EU-15 by choosing this option from the menu `Scope (countries)`. You can use various combinations of search criteria. For example, you can search for all articles that focus on the EU-15, use directive level data, and analyze transposition rather than infringment procedures by selecting the appropriate categories from the menus `Scope (countries)`, `DV aggregation level`, and `DV focus`.
You can also search for all studies that used a certain variable (e.g. `administrative efficiency`) or a certain category of variables (e.g. `institutions`) from the menus `IDV specification` and `IDV category` respectively.
If you do not wish to restrict your results in terms of a particular criteria, you can ignore the relevant menu, or choose `-all-`.
By default, the results page only presents a summary of the information available for each study that fits the search criteria. You can get more details on each study by clicking on the `detials` button to be found at the end of the summary information about the study.
List of variables
Full reference of the study [EJPR style]
Short reference for the analysis [Author (Year) - Roman number].
Each study can contain more than one analysis. Analyses that feature different dependent variables are considered separate analyses. Different models run on the same dependent variable are not considered separate analyses. Results from multinomial regressions are not considered separate analyses.
- N directives
Number of directives on which the analysis is based [NNNNN]
- DV operationalization
Compliance aspect which is used as a dependent variable
Short description of the dependent (outcome) variable. In order to enhance comparability across studies, only few categories are used. The substantive, and not the technical aspect of the dependent variable is taken into account. Thus, even if, technically, the dependent variable in survival analyses is the hazard of transposition, the database reports transposition timeliness as the dependent variable.
- DV precise operationalization
Detailed description of the dependent variable
- DV data
Data source(s) of the dependent variable.
- DV aggregation level
Aggregation level of the dependent variable
The aggregation level can be an article of a directive, a directive, a policy sector, a country, a time period and any combination of these.
- DV focus
Part of the compliance process captured by the dependent variable
This is an indicator for the stage of compliance that is in focus in the particular analysis.
- Scope (countries)
Geographical scope of the analysis
(see below for the list of country abbreviations)
- Scope (time period)
Temporal scope of the analysis [YYYY-YYYY]
- Scope (policy sectors)
The policy sector scope of the analysis
If more than six sectors are covered, the indicator shows ‘all sectors`.
- Method of analysis
Short reference to the statistical method of analysis
- Multi-level controls
Type of controls for the multi-level structure of the analysis used
- Link to dataset
Internet link to the dataset used in the analysis
- IDV specification
Short reference to the independent variable
The independent variable is not always designated exactly as mentioned in the original analysis. In order to ensure comparability across studies, the names of the variables are adjusted slightly while staying as close to the original source as possible. For example, `administrative quality`, `administrative efficiency`, and `bureaucratic quality` are all coded as `administrative efficiency`.
- IDV category
Type of independent variable
This indicator captures the broad category within which the independent variable falls. Again, the classification is done with view to enabling comparison across studies and sometimes the categorization differs from the formulation of the authors of the analysis. Also, this categorization follows closer the operationalization rather than the conceptualization. Main categories:
NIM feature - properties of the national implementing measures (NIM) like number, type, novelty
EU-level conflict - conflict at the EU level for the adoption of the directive
directive feature - properties of the directive like length, type, author, decision rule, etc.
preferences - indicators of preferences taken from surveys of public opinion, expert surveys of party positions, economic proxies, etc.
power - indexes of member states power - most often voting power is used as a proxy but some economic variables are also used
culture - indicators of some aspect of the political (administrative, etc.) culture of a country
institutions - some type of political, administrative, or economic institution
capacity - an indicator of a country`s policy making and/or administrative capacity for policy making and implementation
enforcement - indicators of enforcement activities (like the start of an infringement procedure)
learning - indicators of learning effects, most often capturing the influence of time
other - any other types of variables
- IDV level
Level of variation of the independent variable [directive/country/sector/time]
The level at which the independent variable varies: directive, country, directive/country, sector, time, etc.
- Hypothesized effect
Hypothesized effect of the independent variable
This is a summary of the hypothesized effect of the independent variable. In most cases this is either `positive` or `negative`. More complex hypotheses are summarized to the extent possible (e.g. ‘+/-‘) for a hypothesis of an initially positive effect that turns into a negative one. All hypotheses are expressed in terms of compliance, so the original hypotheses might have been inverted. Hence, a positive sign always means a positive effect on compliance (e.g. shorter transposition delay, shorter time for transposition, lower number of infringement procedures, earlier stage of the infringement procedure reached, etc.) For continuous variables, the effect is expressed for increasing values of the IDV. For example, if the IDV specification reads ‘corporatism` (or ‘political complexity`), the hypothesis is expressed in terms of increasing levels of ‘corporatism` (or ‘political complexity)`. Thus, a positive sign in the Hypothesized effect category means that higher levels of corporatism are hypothesized to have a positive effect on compliance. A negative sign for ‘political complexity` means that increasing political complexity is hypothesized to deteriorate compliance. For binary variables the hypothesis is expressed in terms of the presence of the attribute mentioned in the name of the IDV variable under IDV specification. For example, a positive effect of the variable ‘national package law` means that the presence of national package law increases compliance. For categorical variables, please see the reference category under Operationalization.
Short description of the operationalization of the independent variable
Here you can also find the reference category of binary and categorical variables.
- Operationalization (reference)
Reference to the data source used for measurement of the independent variable
Note: all EU-level indicators that are in principle available from the EURLEX (former CELEX) database are recorded as measured using the EURLEX database, although in fact the researchers might have used a different data source.
- Finding (sign)
The sign of the association between the independent variable and the dependent variable
This is a summary of the estimated association between the IDV and the DV. In most cases, the sign is either positive, negative, or zero. More complex findings are summarized to the extent possible: for example, ‘first + then -` means than an initially positive association turns into a negative one. In general, a positive or a negative sign is reported even for relationships that are not statistically significant. However, sometimes if the relationship is very close to zero (e.g. -0.002), a zero is recorded. All findings are expressed in terms of compliance, so the original findings might have been inverted. Hence, a positive sign always means a positive effect on compliance (e.g. shorter transposition delay, shorter time for transposition, lower number of infringement procedures, earlier stage of the infringement procedure reached, etc.) For continuous variables, the effect is expressed for increasing values of the IDV. For example, if the IDV specification reads ‘corporatism` (or ‘political complexity`), the effect is expressed in terms of increasing levels of ‘Corporatism` (or ‘Political complexity)`. Thus, a positive sign in the Effect (sign) category means that higher levels of corporatism have a positive effect on compliance. A negative sign for ‘political complexity` means that increasing political complexity deteriorates compliance. For binary variables the effects are expressed in terms of the presence of the attribute mentioned in the name of the IDV variable under IDV specification. For example, a positive effect of the variable ‘national package law` means that the presence of national package law increases compliance. For categorical variables, please see the reference category under Operationalization.
- Finding (significance)
Statistical significance of the relationship [yes/no]
Findings significant at the 5% level are reported as significant. If several models with conflicting levels of significance are reported, the interpretation favored in the text of the analysis is taken into account.
- Finding (effect size)
Estimate of the effect size of the relationship between the independent variable and compliance
This category tracks estimates of the effect size of the IDV on compliance. The effect size estimate follows calculations by the authors of the analyses and are not computed independently for the purpose of the database. So, if the authors do not report effect sizes, no records of effect sizes exist in the database even though, in principle, the published information in the analyses is sufficient for the computation of effect sizes.
The readers are advised to take extreme care in interpreting the effect sized because the estimates depend on the precise formulation of the statistical model, the levels at which the other IDV are set, the assumed functional form of the relationship, etc. In addition, the reported effect size depends on the units of measurement of the IDV and the outcome, and on the change in the IDV for which the effect is computed (percentage, standard deviation, point, etc.).
- Finding (effect size - note)
Details about the interpretation of the effect size
Details that enable the interpretation of the effect size, e.g. the type of effect size (odds, hazard ratios, predicted probabilities, etc.) and the change in IDV for which the effect is computed.
List of abbreviations
CPH - Cox proportional hazards
DV - dependent (outcome) variable
IDV - independent variable
LFN - letter of formal notice
NIM - national implementing measure
CEE-8: CZ, EE, HU, LA, LI, PL, SK, SL
EU-27: all member states as of 2007
EU-15: AU, BE, DK, FI, FR, GE, GR, IR, IT, LU, NL, PT, SP, SW, UK
AU - Austria
BE - Belgium
BG - Bulgaria
CY - Cyprus
CZ - Czech Republic
DK - Denmark
EE - Estonia
FI - Finland
FR - France
GE - Germany
GR - Greece
HU - Hungary
IR - Ireland
IT - Italy
LA - Latvia
LI - Lithuania
LU - Luxembourg
MA - Malta
NL - Netherlands
PL - Poland
PT - Portugal
RO - Romania
SK - Slovakia
SL - Slovenia
SP - Spain
SW - Sweden
UK - United Kingdom