Overhead costs that programmes incur in order to operate. In the tool, this is measured as a percentage of programmes' total transfer value. Administrative costs usually range between 5% and 15% of a programme's transfer value, depending on its complexity.

A measure of inequality on scale from 0 to 1, with a higher value indicating higher inequality. It incorporates a inequality aversion parameter, which affects how sensitive this index is to changes at the lower end of the income distribution.

The total annual benefit amount per eligible person. When simulating the transfer, each household receives benefits in monthly installment per eligible person.

A proxy to identify the poorest households by focusing on the bottom 40 per cent (or bottom 4 deciles) in the total household consumption per capita distribution.

A social protection programme that distributes cash benefits to the parents or guardians of children. Child benefit programmes can involve benefits in cash and/or in kind, which can be unconditional or subject to the family complying with conditions such as child's school attendance. The global average benefit level is 4% of GDP per capita.

Classification and regression tree analysis (CART) is an umbrella term for a wide array of supervised machine learning algorithms that help analyze data and visualize results.

Consumption per capita corresponds to total household consumption expenditure divided by number of household members. This variable enters the simulation in monthly frequency.

A numerical value ranging between 0 and 1 that represents the cost of splitting in LNOB analysis. The higher the complexity parameter, the higher the cost of splitting and therefore the smaller or less complex the trees. Conversely, the lower the complexity parameter, the lower the cost of splitting and therefore the larger or more complex the trees.

A social protection programme that distributes cash benefits to persons with disabilities. Depending on the approach adopted, a person can be recognised as having a disability based on impairments, functional limitations, or ability to carry out daily tasks. For the purpose of the simulator, persons with disabilities are defined as individuals living with moderate or severe functional limitations. The global average benefit level is 14% of GDP per capita and the OECD average benefit level is 23% of GDP per capita.

Rules that determine whether individuals or households are eligible to receive benefits. These typically include being below a certain age for child benefits, above a certain age for old age benefits, having proven functional difficulties for disability benefits. They may also include living on less than a certain income (see Means tested programmes) or not being in receipt of benefits from contributory schemes (see Pension tested programmes).

Gross Domestic Product, the value of goods and services produced within a country in a year.

A measure of the degree of inequality in the welfare distribution. It can be conceptualised as ratio of the area between the Lorenz curve (which depicts the percentage of income owned by a given share of the population) and the 45-degree line (which represents perfect equality), and the area below the 45-degree line. The index assumes values from 0 (complete equality) to 1 (complete inequality).

Leaving no one behind

The LNOB analysis is based on CART methodology. Specificually it includes binary regression trees, or LNOB trees for short, based on analysis of variance (ANOVA). There is a two-pronged approach. First, an algorithm selects the circumstances that shape the furthest behind and furthest ahead group, so the results are free of human bias and preconceptions. For each country and each indicator, a unique LNOB tree is produced and therefore the furthest behind in one development area (e.g. stunting in children) are not the same as those in another area (e.g. overweight in children). Second, the CART analysis captures situations when advantages or disadvantages intersect. Frequently, circumstances overlap to create pockets of advantage or disadvantage. Without the CART analysis, revealing these intersections would require multiple calculations and group comparisons to identify who the furthest behind group is in each indicator. LNOB analysis does not have causal interpretation. More rigorous econometric techniques are required to establish causality on a case by case basis. For more details on LNOB analysis, visit ESCAP LNOB Platform here https://lnob.unescap.org/

The visual result of applying LNOB analysis to numerical binary outcome variables such as having access to a basic services represented by 1 for "Yes access" and 0 for "No access". For more details see, LNOB analysis.

Programmes that require recipients to be living below a certain level of income or wealth.

A numerical value representing the minimum number of observations or share of the sample in a terminal node. This is a stopping criterion to ensure that the disaggregation does not continue when the number of observations in a group gets “too small” to take an average, as this could lead to biased and/or inconsistent estimates. In practice, at least a minimum of 100 observations are required in terminal nodels.

A social protection programme that distributes cash benefits to persons above retirement age (which may vary across countries). Also referred to as 'social pension', not to be confused with contributory pension schemes. The global average benefit level is 16% of GDP per capita and the OECD average benefit level is 22% of GDP per capita.

A sankey diagram is a visualization used to depict a flow from one set of values to another.

According to the ILO, social protection (or social security) is understood as a collection of policies and measures that reduce income insecurities and prevent vulnerabilities across the lifecycle. A social protection system should ensure that all individuals during the course of their life, at the very least, have an adequate standard of living. Social protection schemes should reduce and prevent poverty and vulnerability, as well as provide a buffer against contingencies across the lifecycle and environmental shocks. Typically, social protection programmes include child or family benefits, disability benefits, unemployment benefits and old age pensions

Social Protection Online Toolbox

Social Protection to Leave No One Behind

A measure of inequality as part of the Generalised Entropy measures. Value 0 indicates perfect ineuqality and a higher value indicatoes higher inequality.