DIAfragma

Municipality of Amsterdam

The DIAfragma creates an integrated customer picture of minimum households for the Poverty Reduction Department.

Last change on 10th of September 2024, at 14:10 (CET) | Publication Standard 1.0

• Publication category: Impactful algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Social Security

Begin date

Field not filled in.

Contact information

[e-mail]

Responsible use

Goal and impact

The Poverty Alleviation Department wants everyone eligible for low-income assistance schemes to make use of them. Therefore, we want to approach potential beneficiaries proactively. To map this group, we use various sources. We call this dataset of data the DIAfragma. The DIAfragma creates an integral customer picture of minimum households for the Poverty Reduction Department. This provides up-to-date information on the total client base and allows us to proactively approach the potential minimum target group.

Considerations

DiaFragma only works with BSN numbers without further linking with data that could lead to (unjustified) discrimination.

We do not process any special personal data in the sense of Article 9 AVG. However, we do process data such as gender, age and family relationships.

Gender is necessary because we link different systems in which gender is recorded. It is not technically possible not to include gender in this linkage.

Age is needed because age is a requirement for the allocation of certain minimum schemes. Think of child credit based on the age of the child or free public transport for AOW recipients.

Family relationships are needed because we use this data to determine whether someone is entitled to mini benefits, or because the mini benefit is only for a family.

Human intervention

In the 'secure minima' group, we do a human check on a random basis.

In the group of 'uncertain minima', the potential customer receives a letter requesting an (abridged) application. The abridged application form is attached to the letter and contains information that we already know about the potential customer. We ask the potential client to check it, amend it where necessary, sign it and then send it in. Staff from the Poverty Reduction Department process the abbreviated applications . We handle them in a similar way to an application from any citizen. We have established work processes for processing an application.

Risk management

Risk of faulty data:

Error in loading data into the DIAfragm which means that not all (potential) candidates are included and therefore we miss (potential) candidates.

Mitigating measure:

Loading data from internal sources (ZorgNed, Socrates) is done daily. We monitor the result of this loading action daily. If necessary, we take immediate action to fix it.

Loading the external sources is done manually.

Risk of data selection error:

Inappropriate allocation of schemes due to storing personal data for too long.

Mitigating measure:

Make a number of people responsible for maintaining the retention periods of the lists we use to write to (potential) minima.

risk factors:

Personal data (personal data)

Large scale data processing (large-scale data processing)

Combining datasets

Research purposes

Sensitive data

Vulnerable groups

Automated decisions

By risk:

'Secure' and 'uncertain' minima is a group of about 124,000 people. The potential risks apply to the whole group.

Mitigation:

Loading of data from internal sources (ZorgNed, Socrates) is done daily. We monitor the result of this loading action daily. If necessary, we take immediate action to fix it.

Loading the external sources is done manually.

Legal basis

Public interest or public authority

Elaboration on impact assessments

Doing so in response to the commissioning of DIAfragma. The DPIA is now a draft. As soon as it is ready, it will be published.

Operations

Data

The algorithm uses this dataset:

Just

• Name and surname
• Address
• Postal code
• City
• Date of birth
• E-mail address
• Phone number
• Gender
• Family composition
• Indication primary or secondary education (which type of education does a child in the family follow)

Sensitive

• BSN
• IBAN
• Income details
• Indication Supplementary Income Support for the Elderly (yes/no)
• Tax relief data (indication yes/no)

We do not process any special or criminal personal data.

This is a simple set of calculation rules that uses data analysis to map the (potentially) entitled population ('certain' and 'uncertain' minima).

Technical design

We use data from various source files.

Internal sources:

• Living allowances (Socrates via DWH*)
• Special assistance benefits (Socrates via DWH*)
• Poverty benefits (ZorgNed via DWH*)
• Assessed incomes (Socrates, ZorgNed via DWH*)
• Households and persons (Socrates, ZorgNed via DWH*)

* DWH stands for "DataWareHouse WPI".

External sources:

• Name and surname, Address, Postcode and Place of residence (Municipality of Amsterdam, Basisregistratie Personen (BRP), Directorate of Data). Note! Only required for newly or re-identified minima) who receive the general leaflet on poverty benefits.
• BSN (Municipality of Amsterdam, Basisregistratie Personen (BRP), Directorate Data. Attention! Only needed for newly or re-identified minima) who receive an ex officio allocation and/or abbreviated application form, as far as facilities are concerned where the household composition is important, to determine with whom the household is shared and they can also be notified.
• Supplementary Income Support for the Elderly (AIO, SVB via portal Inlichtingenbureau - indication)
• Discharges (Municipality of Amsterdam, Directorate of Taxes - indication)
• Children in primary or secondary education (Municipality of Amsterdam, Directorate OJZD - indication).

The DIAfragma is a dataset of data off from WPI's various source systems. With this, we create an up-to-date customer view . Such a customer view contains information such as name, date of birth, household and level of income (in % of the social minimum) and source of income. This means keeping the information available to the Poverty Reduction Department up to date and at a high level of quality. Available information comes, among other things, from requests by Amsterdam residents, giving the department permission to see income data and limited asset data (house and car ownership). Household information is obtained by the department from the BRP. We request income data of self-employed people from the client via a letter containing a request for information.

From various sources (data from source systems in the WPI DWH and possible external sources, we select 'certain' and 'uncertain' minimum clients....

To enrich the BRP data, we add citizen service numbers (BSNs).

The enriched return file contains the complete household: the client with any partner and any children.

We compare this file with the households known in our DWH. Based on this, we determine an optimal household. So that missing data are now completed. For example, a child that was not yet known in our systems but is known in BRP etc.