Rationale – RIP Counties

For much of 2020, and especially during our lockdowns, the Irish mainstream media and public officials repeatedly announced death figures – while RIP.ie was showing things were fairly normal in most places.

In stark contrast, the very same media and officials have since grown silent on death figures – while RIP.ie has begun showing some unusual increases in the numbers dying.

Why are the media not reporting on this?

If you ask the journalists, editors and owners directly, you may get a surprise: they don’t have to cover it.

As Patrick E. Walshe learned, when he complained to the Broadcasting Authority of Ireland, the licenced media in Ireland are under no obligation to report or even investigate the rising death numbers.

What about official sources instead?

Official sources can take months, quarters or years to release statistical data or reports. While this provides a benefit in terms of accuracy and thoroughness, it can mean missing signals in the data as they emerge in “real time”.

Furthermore, methods of analysis may be unclear and the raw data itself might not be easily accessible or might even be deliberately withheld.¹ This makes it harder for others to verify the work and any resulting conclusions or claims.

In contrast RIP.ie provides a publicly available record of deaths which is updated daily.

Furthermore, we tell you:

how we gathered data from it and
how we generated our graphs and tables.

So you can verify our work and ‘do your own county’ if you wish.

But it’s grass-roots. Can we trust RIP.ie data?

Yes.

According to Central Statistics Office analyses:

“Comparing CSO data with RIP.ie death notices for 2016 and 2017 showed a strong correlation of over 99%. This indicates that the number of death notices on RIP.ie can be used to accurately indicate trends in mortality…

… RIP.ie is a privately-owned company and over the last decade or so has become a popular website to notify deaths and provide funeral details, both for Irish and Northern Irish citizens and expatriates. Funeral directors have a secure login and can post a notice without cost on the website. Only funeral directors can post a notice, ensuring the data is of good quality and duplicates are rare. Due to the Irish custom of holding funerals within two to three days following death, these notices are usually placed in a fast and efficient manner. The combination of these market and cultural forces have made RIP.ie a valuable crowd-sourced means of tracking deaths. These notices are close to ‘real time’ – the average length of time between date of death and publication is about 1.1 days. When compared with the statutory time limit of three months for the registrations of deaths in the State, this is a timelier data source for monitoring trends in mortality…
Measuring Mortality Using Public Data Sources
October 2019 – June 2020: Experimental Analysis
CSO statistical release, 03 July 2020, 11am²

How far back?

“… Before 2017, the coverage of deaths on RIP.ie was not as comprehensive as in recent years and we observe that usage of RIP.ie steadily decreases as one moves backward in time.³ For this reason, if we were to use the RIP.ie data before 2017, it would increase the estimate of excess mortality for 2020/2021.

Annual deaths are increasing over the last decade due in part at least to an ageing population. It is important to differentiate between this natural demographic effect and the effect of the pandemic. This is another reason to look at more recent years to establish expected mortality
Measuring Mortality Using Public Data Sources 2019-2021
October 2019 – March 2021: Experimental Analysis
CSO statistical release, 30 April 2021, 11am⁴

However, comparing the number of RIP.ie records for a county with its overall population is another way we might spot changes in death patterns. So, as a census was taken in 2016 (providing us with county population figures), our contributors included that year when gathering data from RIP.ie

It was also decided, to help readers to compare our results with others⁵ to include the year 2015.

(Note: some early contributions to this project were completed before the suggested 2016 cut-off point was discovered. So some archived counties contain data from earlier years).

There are caveats though

Some deaths on RIP.ie are duplicated – for legitimate, well-intentioned reasons.⁶

As well as ocurring within a county, duplication can also cross county lines.⁷

Furthermore, sometimes deaths that occur abroad are recorded⁸ – and those deaths can also be duplicated on occasion.⁹

Also a person’s age, sex or cause of death are usually not given.¹⁰ So RIP.ie records are not going to help answer specific questions like:

are cardiac-related deaths on the rise?
are more 30-39 or 40-49 year olds dying now than usual?
are men/boys dying at a greater rate than women/girls?
were more life-years lost in 2020 than 2019?¹¹ Etc.

Why bother with RIP.ie data then?

Because they can show us trends in all-cause mortality. Which enables a view of ‘the big picture’ without getting lost in detailed debates about why a person died.

“… deaths can be neither hidden nor manipulated. It’s binary. One or zero. Either he died or he didn’t…

Meanwhile, making claims about the cause of death quickly becomes complicated due to (lack of) supporting evidence and other epistemological challenges.

For example, if Mike told me that Jenny died from Covid, then my response might be “how do you know?” At which point, Mike is compelled to provide supporting evidence. If he responds with something like “the doctor said so”, then my response might be “how does the doctor know?”.¹²

It becomes messy.
Jerm Warfa r e

Furthermore, as mentioned above, unlike official data RIP.ie data is available in near real-time

Cleaning the data

It is unclear but seems reasonable to assume that the Central Statistics Office formally obtained a detailed spreadsheet of all RIP.ie records. This would have made cleaning the data fairly straightforward.

In contrast those ‘doing our own counties’ could only access data via the public search form. This meant we were copying the headlines of 40 records at a time and pasting them into Excel or LibreOffice Calc documents… before clicking ‘next’ and copying the next page of 40 headlines… then the next 40… then the next 40… etc

After tidying the formatting¹³ we searched for duplicate names (using functions built into the spreadsheet software) and deleted them.

However, finding and removing deaths that occurred abroad was deemed prohibitively difficult. Because to do so would have involved opening each record manually and reading it, before deciding on its removal. This is perhaps manageable over a few dozen records… but not many thousands.¹⁴ So, contributors did not search for or remove individual deaths which occurred abroad. But duplicates of their notices were removed when found within a county.

For the same reason our contributors did not search beyond a county for duplicates occurring in another county. Or Irish language duplicates of English names.

Are those limitations grounds for dismissal?

If we were trying for a definitive report, then yes. But we are not trying to be definitive. We are simply taking a ‘broad-strokes’ look at RIP.ie for signals that might suggest there’s something going on that needs further conversation and investigation.

So, for that purpose, we believe our approach is valid and will suffice.

Comparing each county with itself

Research assumptions are important. Declaring them is too. For example…

… imagine if a study claimed 100% of participants who leapt from an aeroplane without a parachute were unharmed.

Unbelievable? Impossible? Yes. If you assume the aeroplane is in flight!

But what if the researchers…

assumed there was no need to research the effects of jumping from a flying aeroplane, but that a stationary aeroplane would suffice
and didn’t tell you they had made that assumption?!

So, we are declaring our key assumptions

(Note: as outlined here, we are now removing duplicates from national level data before tallying each county. Which makes the following section irrelevant. But we’re leaving it in as it may assist others doing their own county.

For the most part, those posting on RIP.ie would only be posting for their own locality, their own county. But even if they also posted a death in a neighbouring county, we do not need to track down that duplicate. Because, as the CSO remarked, duplicates are rare and we are not studying the entire country nor comparing death rates in one county to those in another.

Instead we are assuming that, if surveyed over a few years, a picture would emerge of the ‘normal’ death patterns for an individual county, giving us a sense of…

the numbers dying in that county from year to year
the percentage of its population that die from year to year, and
its seasonal death patterns¹⁵…

… thereby creating a backdrop against which sudden changes or anomalies might become apparent…

1
Two examples from the USA. (1) It took 464 days and two court cases before the Informed Consent Action Network (ICAN) could access the anonymised data on Covid vaccine side-effects collected by the CDC through their purpose-built “V-safe After Vaccination Health Checker smartphone app. 10+million vaccine recipients voluntarily signed up for the app. Of them 33% reported side-effects, 25% required time off work or school and 7.7% required medical attention. But the CDC resisted letting that be publicly known and as of April 2023 have still yet to release all of what those 10+million volunteers reported. Links to lawyer Aaron Siri’s story here. (2) The FDA sought to delay for 75 years, the releasing of Pfizer’s post-marketing Covid-19 vaccine safety data (when sued by the PHMPT). But a court ruled that they could have just 8 months delay instead. There’s a video overview of the ‘Pfizer Documents’ by Sonjia Elijah (TrialSite News) here. Thorough analysis of the 677 documents can be found on DailyClout.io (summary of findings and DailyClout’s foundation here).
2
Archive
3
Gerry O’Neill touched on this issue in Do Sligo Lives Matter?
4
Archive
5
For example: Ireland Excess Deaths and Irish Quislings
6
For instance, “John Kelly” might have lived his first twenty years in Casla, Co. Galway then moved to Galway city where he died fifty years later. Someone who knew him in Galway as John Kelly might submit a notice. Meanwhile two people who knew of his Casla background – and want to ensure people in Casla know of his passing in Galway (because that’s how we Irish roll) – might each submit separate notices about a John Kelly and a Seán Ó’Ceallaigh (because Casla is in The Gaeltacht where Irish is the primary language). So his death may be listed three times in Co. Galway..
7
If “John Kelly” from our previous example had instead moved to live in Castlebar, his death might appear in Co. Mayo as well as in Co. Galway.
8
e.g. Northern Ireland, England, USA…
9
Let’s say a Leitrim native died in New York. A notice might be placed (a) close to the day they died and (b) a month or more later announcing ceremonies in Leitrim for interring their body or ashes locally, or simply to commemorate them.
10
While they can sometimes be gleaned manually from some records, the inconsistent recording of these characteristics is limiting.
11
Putting it crudely… if average life expectancy was 80 years of age, a death of a 30 year-old would count as 50 life-years lost. So, at a population level, an increase in life-years lost over the yearly average would indicate something was adversely affecting younger cohorts. This Joel Smalley article gives insight into the usefulness of the life-years lost metric.
12
This raises the topic of Covid testing. Issues with the use of PCR for this purpose were raised by the Corman-Drosten Review Group in November 2020. See also this historical perspective by Dr. Thomas Binder, a member of that group
13
e.g. misaligned columns, unwanted column headers
14
Even a lightly populated county like Leitrim was seeing c. 400 RIP.ie records per year (until jumping to 501 in 2022). So going back to 2016 means c. 2,800 records. Or Galway… with c 2,000 per year (c. 2,400 in 2022) means 14,000 records to individually click and read.
15
Usually greater numbers die in Winter than Summer