Understanding the migrant death issue
The data on migrant deaths comes from the Arizona OpenGIS Project, cosponsored by Humane Borders and the Pima County Office of Medical Examiner. I’ve edited the data set to include all recorded migrant deaths between 2000 up through December of 2026. My goal with any analysis of these data is to always honor and humanize those who have died in the desert. These data permit a better understanding of the death crisis. This report disaggregates migrant remains for each month for all calendar years.
A note about the data
The migrant death crisis effectively “begins” around the year 2000. The reason the crisis was nonexistent before 2000 is that very, very few people crossed through Arizona. This changed due to changes made in the Clinton Administration that led to the funnel effect pushing migrants into the Arizona corridor. So if one downloads the death map data, one will find that 4,505 migrant remains have been recovered (as of 7/31/2026); however, 4,371 have been recovered since the year 2000. This means that in the death map, only 134 remains are recorded as being found between 1981 and 1999 (an 18 year period). I don’t mean “only” in a demeaning way; rather, relative to the mass death crisis, the total number of deaths reported in the 18 year period between 1981 to 1999 is far lower than the average yearly number of deaths after the year 2000.
The data file is a csv file saved to my GitHub site. I dynamically update the data as new information is added. This file is current through 7/31/26.
Migrant deaths by month and year
The following analysis begins with January and culminates with December. The analysis for each month is identical: number of remains; number of remains sorted from lowest to highest; number of remains accounting for PMI in two ways. For each month, there are four plots.
January
Table of remains recovered
| January_data$year_month | n | percent |
|---|---|---|
| Jan 2000 | 4 | 1.53% |
| Jan 2001 | 1 | 0.38% |
| Jan 2002 | 3 | 1.15% |
| Jan 2003 | 6 | 2.30% |
| Jan 2004 | 7 | 2.68% |
| Jan 2005 | 5 | 1.92% |
| Jan 2006 | 4 | 1.53% |
| Jan 2007 | 18 | 6.90% |
| Jan 2008 | 4 | 1.53% |
| Jan 2009 | 7 | 2.68% |
| Jan 2010 | 15 | 5.75% |
| Jan 2011 | 5 | 1.92% |
| Jan 2012 | 13 | 4.98% |
| Jan 2013 | 7 | 2.68% |
| Jan 2014 | 13 | 4.98% |
| Jan 2015 | 10 | 3.83% |
| Jan 2016 | 8 | 3.07% |
| Jan 2017 | 13 | 4.98% |
| Jan 2018 | 6 | 2.30% |
| Jan 2019 | 14 | 5.36% |
| Jan 2020 | 17 | 6.51% |
| Jan 2021 | 20 | 7.66% |
| Jan 2022 | 15 | 5.75% |
| Jan 2023 | 6 | 2.30% |
| Jan 2024 | 15 | 5.75% |
| Jan 2025 | 12 | 4.60% |
| Jan 2026 | 13 | 4.98% |
| Total | 261 | - |
Visualizing migrant deaths by January and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
January_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 49 | 19% | 21% |
| < 1 week | 3 | 1% | 1% |
| < 3 months | 24 | 9% | 10% |
| < 3 weeks | 3 | 1% | 1% |
| < 5 weeks | 8 | 3% | 3% |
| < 6-8 months | 42 | 16% | 18% |
| > 6-8 months | 105 | 40% | 45% |
| NA | 27 | 10% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Jan 2000 | 25% (1) | 0% (0) | 0% (0) | 0% (0) | 25% (1) | 25% (1) | 25% (1) | 0% (0) | 100% (4) |
| Jan 2001 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (1) | 0% (0) | 100% (1) |
| Jan 2002 | 67% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 33% (1) | 0% (0) | 100% (3) |
| Jan 2003 | 33% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 17% (1) | 33% (2) | 17% (1) | 100% (6) |
| Jan 2004 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (7) | 0% (0) | 100% (7) |
| Jan 2005 | 80% (4) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 0% (0) | 100% (5) |
| Jan 2006 | 50% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 25% (1) | 25% (1) | 0% (0) | 100% (4) |
| Jan 2007 | 39% (7) | 0% (0) | 0% (0) | 0% (0) | 11% (2) | 11% (2) | 39% (7) | 0% (0) | 100% (18) |
| Jan 2008 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 25% (1) | 75% (3) | 0% (0) | 100% (4) |
| Jan 2009 | 14% (1) | 0% (0) | 0% (0) | 29% (2) | 0% (0) | 0% (0) | 57% (4) | 0% (0) | 100% (7) |
| Jan 2010 | 47% (7) | 7% (1) | 7% (1) | 0% (0) | 0% (0) | 13% (2) | 20% (3) | 7% (1) | 100% (15) |
| Jan 2011 | 60% (3) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 20% (1) | 100% (5) |
| Jan 2012 | 8% (1) | 0% (0) | 31% (4) | 0% (0) | 0% (0) | 31% (4) | 15% (2) | 15% (2) | 100% (13) |
| Jan 2013 | 0% (0) | 0% (0) | 43% (3) | 14% (1) | 0% (0) | 14% (1) | 29% (2) | 0% (0) | 100% (7) |
| Jan 2014 | 8% (1) | 8% (1) | 8% (1) | 0% (0) | 0% (0) | 46% (6) | 23% (3) | 8% (1) | 100% (13) |
| Jan 2015 | 10% (1) | 0% (0) | 10% (1) | 0% (0) | 0% (0) | 50% (5) | 10% (1) | 20% (2) | 100% (10) |
| Jan 2016 | 13% (1) | 0% (0) | 13% (1) | 0% (0) | 13% (1) | 25% (2) | 25% (2) | 13% (1) | 100% (8) |
| Jan 2017 | 0% (0) | 0% (0) | 15% (2) | 0% (0) | 8% (1) | 38% (5) | 38% (5) | 0% (0) | 100% (13) |
| Jan 2018 | 0% (0) | 0% (0) | 17% (1) | 0% (0) | 0% (0) | 33% (2) | 33% (2) | 17% (1) | 100% (6) |
| Jan 2019 | 0% (0) | 7% (1) | 21% (3) | 0% (0) | 0% (0) | 7% (1) | 43% (6) | 21% (3) | 100% (14) |
| Jan 2020 | 24% (4) | 0% (0) | 12% (2) | 0% (0) | 0% (0) | 12% (2) | 53% (9) | 0% (0) | 100% (17) |
| Jan 2021 | 25% (5) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 55% (11) | 20% (4) | 100% (20) |
| Jan 2022 | 7% (1) | 0% (0) | 13% (2) | 0% (0) | 13% (2) | 20% (3) | 47% (7) | 0% (0) | 100% (15) |
| Jan 2023 | 33% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 17% (1) | 33% (2) | 17% (1) | 100% (6) |
| Jan 2024 | 7% (1) | 0% (0) | 0% (0) | 0% (0) | 7% (1) | 7% (1) | 47% (7) | 33% (5) | 100% (15) |
| Jan 2025 | 17% (2) | 0% (0) | 25% (3) | 0% (0) | 0% (0) | 0% (0) | 33% (4) | 25% (3) | 100% (12) |
| Jan 2026 | 8% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 8% (1) | 77% (10) | 8% (1) | 100% (13) |
| Total | 21% (49) | 1% (3) | 8% (24) | 2% (3) | 3% (8) | 16% (42) | 41% (105) | 9% (27) | 100% (261) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
February
Table of remains recovered
| February_data$year_month | n | percent |
|---|---|---|
| Feb 2000 | 7 | 2.49% |
| Feb 2001 | 2 | 0.71% |
| Feb 2002 | 4 | 1.42% |
| Feb 2003 | 10 | 3.56% |
| Feb 2004 | 12 | 4.27% |
| Feb 2005 | 7 | 2.49% |
| Feb 2006 | 5 | 1.78% |
| Feb 2007 | 13 | 4.63% |
| Feb 2008 | 11 | 3.91% |
| Feb 2009 | 8 | 2.85% |
| Feb 2010 | 16 | 5.69% |
| Feb 2011 | 15 | 5.34% |
| Feb 2012 | 15 | 5.34% |
| Feb 2013 | 11 | 3.91% |
| Feb 2014 | 8 | 2.85% |
| Feb 2015 | 6 | 2.14% |
| Feb 2016 | 1 | 0.36% |
| Feb 2017 | 15 | 5.34% |
| Feb 2018 | 10 | 3.56% |
| Feb 2019 | 11 | 3.91% |
| Feb 2020 | 21 | 7.47% |
| Feb 2021 | 15 | 5.34% |
| Feb 2022 | 11 | 3.91% |
| Feb 2023 | 15 | 5.34% |
| Feb 2024 | 18 | 6.41% |
| Feb 2025 | 8 | 2.85% |
| Feb 2026 | 6 | 2.14% |
| Total | 281 | - |
Visualizing migrant deaths by February and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
February_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 67 | 24% | 26% |
| < 1 week | 11 | 4% | 4% |
| < 3 months | 19 | 7% | 7% |
| < 3 weeks | 2 | 1% | 1% |
| < 5 weeks | 1 | 0% | 0% |
| < 6-8 months | 29 | 10% | 11% |
| > 6-8 months | 131 | 47% | 50% |
| NA | 21 | 7% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Feb 2000 | 86% (6) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 0% (0) | 100% (7) |
| Feb 2001 | 0% (0) | 0% (0) | 50% (1) | 0% (0) | 0% (0) | 0% (0) | 50% (1) | 0% (0) | 100% (2) |
| Feb 2002 | 50% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (2) | 0% (0) | 100% (4) |
| Feb 2003 | 20% (2) | 0% (0) | 10% (1) | 0% (0) | 0% (0) | 0% (0) | 70% (7) | 0% (0) | 100% (10) |
| Feb 2004 | 50% (6) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 8% (1) | 17% (2) | 25% (3) | 100% (12) |
| Feb 2005 | 86% (6) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 0% (0) | 100% (7) |
| Feb 2006 | 40% (2) | 20% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 40% (2) | 0% (0) | 100% (5) |
| Feb 2007 | 46% (6) | 15% (2) | 0% (0) | 0% (0) | 0% (0) | 15% (2) | 0% (0) | 23% (3) | 100% (13) |
| Feb 2008 | 36% (4) | 9% (1) | 0% (0) | 0% (0) | 0% (0) | 9% (1) | 36% (4) | 9% (1) | 100% (11) |
| Feb 2009 | 38% (3) | 13% (1) | 0% (0) | 0% (0) | 13% (1) | 13% (1) | 25% (2) | 0% (0) | 100% (8) |
| Feb 2010 | 44% (7) | 19% (3) | 13% (2) | 0% (0) | 0% (0) | 6% (1) | 19% (3) | 0% (0) | 100% (16) |
| Feb 2011 | 7% (1) | 7% (1) | 7% (1) | 13% (2) | 0% (0) | 20% (3) | 47% (7) | 0% (0) | 100% (15) |
| Feb 2012 | 13% (2) | 7% (1) | 7% (1) | 0% (0) | 0% (0) | 20% (3) | 40% (6) | 13% (2) | 100% (15) |
| Feb 2013 | 18% (2) | 0% (0) | 18% (2) | 0% (0) | 0% (0) | 18% (2) | 45% (5) | 0% (0) | 100% (11) |
| Feb 2014 | 0% (0) | 0% (0) | 25% (2) | 0% (0) | 0% (0) | 38% (3) | 25% (2) | 13% (1) | 100% (8) |
| Feb 2015 | 17% (1) | 0% (0) | 17% (1) | 0% (0) | 0% (0) | 50% (3) | 17% (1) | 0% (0) | 100% (6) |
| Feb 2016 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (1) | 0% (0) | 100% (1) |
| Feb 2017 | 7% (1) | 0% (0) | 7% (1) | 0% (0) | 0% (0) | 20% (3) | 67% (10) | 0% (0) | 100% (15) |
| Feb 2018 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 70% (7) | 30% (3) | 100% (10) |
| Feb 2019 | 0% (0) | 0% (0) | 27% (3) | 0% (0) | 0% (0) | 0% (0) | 45% (5) | 27% (3) | 100% (11) |
| Feb 2020 | 5% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 10% (2) | 76% (16) | 10% (2) | 100% (21) |
| Feb 2021 | 0% (0) | 7% (1) | 13% (2) | 0% (0) | 0% (0) | 13% (2) | 60% (9) | 7% (1) | 100% (15) |
| Feb 2022 | 27% (3) | 0% (0) | 9% (1) | 0% (0) | 0% (0) | 9% (1) | 45% (5) | 9% (1) | 100% (11) |
| Feb 2023 | 53% (8) | 0% (0) | 7% (1) | 0% (0) | 0% (0) | 0% (0) | 40% (6) | 0% (0) | 100% (15) |
| Feb 2024 | 22% (4) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 6% (1) | 72% (13) | 0% (0) | 100% (18) |
| Feb 2025 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (8) | 0% (0) | 100% (8) |
| Feb 2026 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 83% (5) | 17% (1) | 100% (6) |
| Total | 25% (67) | 4% (11) | 8% (19) | 0% (2) | 0% (1) | 9% (29) | 47% (131) | 7% (21) | 100% (281) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
March
Table of remains recovered
| March_data$year_month | n | percent |
|---|---|---|
| Mar 2000 | 6 | 2.30% |
| Mar 2002 | 3 | 1.15% |
| Mar 2003 | 3 | 1.15% |
| Mar 2004 | 14 | 5.36% |
| Mar 2005 | 7 | 2.68% |
| Mar 2006 | 20 | 7.66% |
| Mar 2007 | 11 | 4.21% |
| Mar 2008 | 3 | 1.15% |
| Mar 2009 | 17 | 6.51% |
| Mar 2010 | 10 | 3.83% |
| Mar 2011 | 15 | 5.75% |
| Mar 2012 | 8 | 3.07% |
| Mar 2013 | 12 | 4.60% |
| Mar 2014 | 11 | 4.21% |
| Mar 2015 | 6 | 2.30% |
| Mar 2016 | 8 | 3.07% |
| Mar 2017 | 6 | 2.30% |
| Mar 2018 | 9 | 3.45% |
| Mar 2019 | 8 | 3.07% |
| Mar 2020 | 13 | 4.98% |
| Mar 2021 | 20 | 7.66% |
| Mar 2022 | 18 | 6.90% |
| Mar 2023 | 15 | 5.75% |
| Mar 2024 | 3 | 1.15% |
| Mar 2025 | 6 | 2.30% |
| Mar 2026 | 9 | 3.45% |
| Total | 261 | - |
Visualizing migrant deaths by March and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
March_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 78 | 30% | 32% |
| < 1 week | 9 | 3% | 4% |
| < 3 months | 18 | 7% | 7% |
| < 3 weeks | 7 | 3% | 3% |
| < 5 weeks | 2 | 1% | 1% |
| < 6-8 months | 34 | 13% | 14% |
| > 6-8 months | 99 | 38% | 40% |
| NA | 14 | 5% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Mar 2000 | 67% (4) | 33% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (6) |
| Mar 2002 | 100% (3) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (3) |
| Mar 2003 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 67% (2) | 33% (1) | 100% (3) |
| Mar 2004 | 71% (10) | 7% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (2) | 7% (1) | 100% (14) |
| Mar 2005 | 100% (7) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (7) |
| Mar 2006 | 60% (12) | 5% (1) | 5% (1) | 0% (0) | 0% (0) | 5% (1) | 20% (4) | 5% (1) | 100% (20) |
| Mar 2007 | 55% (6) | 0% (0) | 9% (1) | 0% (0) | 0% (0) | 0% (0) | 36% (4) | 0% (0) | 100% (11) |
| Mar 2008 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (3) | 0% (0) | 100% (3) |
| Mar 2009 | 47% (8) | 6% (1) | 0% (0) | 12% (2) | 0% (0) | 6% (1) | 18% (3) | 12% (2) | 100% (17) |
| Mar 2010 | 20% (2) | 0% (0) | 0% (0) | 0% (0) | 10% (1) | 30% (3) | 40% (4) | 0% (0) | 100% (10) |
| Mar 2011 | 20% (3) | 0% (0) | 7% (1) | 7% (1) | 0% (0) | 13% (2) | 47% (7) | 7% (1) | 100% (15) |
| Mar 2012 | 0% (0) | 13% (1) | 13% (1) | 0% (0) | 0% (0) | 25% (2) | 38% (3) | 13% (1) | 100% (8) |
| Mar 2013 | 25% (3) | 8% (1) | 25% (3) | 0% (0) | 0% (0) | 25% (3) | 17% (2) | 0% (0) | 100% (12) |
| Mar 2014 | 18% (2) | 0% (0) | 9% (1) | 0% (0) | 0% (0) | 55% (6) | 9% (1) | 9% (1) | 100% (11) |
| Mar 2015 | 33% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 67% (4) | 0% (0) | 100% (6) |
| Mar 2016 | 13% (1) | 0% (0) | 38% (3) | 25% (2) | 0% (0) | 0% (0) | 25% (2) | 0% (0) | 100% (8) |
| Mar 2017 | 0% (0) | 0% (0) | 17% (1) | 0% (0) | 0% (0) | 0% (0) | 67% (4) | 17% (1) | 100% (6) |
| Mar 2018 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 56% (5) | 33% (3) | 11% (1) | 100% (9) |
| Mar 2019 | 0% (0) | 0% (0) | 0% (0) | 13% (1) | 13% (1) | 13% (1) | 50% (4) | 13% (1) | 100% (8) |
| Mar 2020 | 15% (2) | 0% (0) | 15% (2) | 0% (0) | 0% (0) | 8% (1) | 54% (7) | 8% (1) | 100% (13) |
| Mar 2021 | 10% (2) | 5% (1) | 0% (0) | 5% (1) | 0% (0) | 20% (4) | 55% (11) | 5% (1) | 100% (20) |
| Mar 2022 | 28% (5) | 6% (1) | 17% (3) | 0% (0) | 0% (0) | 11% (2) | 39% (7) | 0% (0) | 100% (18) |
| Mar 2023 | 27% (4) | 0% (0) | 7% (1) | 0% (0) | 0% (0) | 13% (2) | 53% (8) | 0% (0) | 100% (15) |
| Mar 2024 | 33% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 67% (2) | 0% (0) | 100% (3) |
| Mar 2025 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 17% (1) | 83% (5) | 0% (0) | 100% (6) |
| Mar 2026 | 11% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 78% (7) | 11% (1) | 100% (9) |
| Total | 29% (78) | 3% (9) | 6% (18) | 2% (7) | 1% (2) | 11% (34) | 41% (99) | 6% (14) | 100% (261) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
April
Table of remains recovered
| April_data$year_month | n | percent |
|---|---|---|
| Apr 2000 | 5 | 1.79% |
| Apr 2001 | 2 | 0.72% |
| Apr 2002 | 7 | 2.51% |
| Apr 2003 | 7 | 2.51% |
| Apr 2004 | 16 | 5.73% |
| Apr 2005 | 9 | 3.23% |
| Apr 2006 | 17 | 6.09% |
| Apr 2007 | 14 | 5.02% |
| Apr 2008 | 17 | 6.09% |
| Apr 2009 | 10 | 3.58% |
| Apr 2010 | 14 | 5.02% |
| Apr 2011 | 12 | 4.30% |
| Apr 2012 | 11 | 3.94% |
| Apr 2013 | 14 | 5.02% |
| Apr 2014 | 13 | 4.66% |
| Apr 2015 | 7 | 2.51% |
| Apr 2016 | 8 | 2.87% |
| Apr 2017 | 6 | 2.15% |
| Apr 2018 | 10 | 3.58% |
| Apr 2019 | 9 | 3.23% |
| Apr 2020 | 10 | 3.58% |
| Apr 2021 | 13 | 4.66% |
| Apr 2022 | 13 | 4.66% |
| Apr 2023 | 10 | 3.58% |
| Apr 2024 | 9 | 3.23% |
| Apr 2025 | 8 | 2.87% |
| Apr 2026 | 8 | 2.87% |
| Total | 279 | - |
Visualizing migrant deaths by April and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
April_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 95 | 34% | 36% |
| < 1 week | 21 | 8% | 8% |
| < 3 months | 12 | 4% | 5% |
| < 3 weeks | 15 | 5% | 6% |
| < 5 weeks | 7 | 3% | 3% |
| < 6-8 months | 28 | 10% | 11% |
| > 6-8 months | 84 | 30% | 32% |
| NA | 17 | 6% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Apr 2000 | 80% (4) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 0% (0) | 100% (5) |
| Apr 2001 | 50% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (1) | 0% (0) | 100% (2) |
| Apr 2002 | 71% (5) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 0% (0) | 14% (1) | 0% (0) | 100% (7) |
| Apr 2003 | 43% (3) | 14% (1) | 14% (1) | 0% (0) | 0% (0) | 14% (1) | 14% (1) | 0% (0) | 100% (7) |
| Apr 2004 | 75% (12) | 0% (0) | 0% (0) | 13% (2) | 0% (0) | 0% (0) | 13% (2) | 0% (0) | 100% (16) |
| Apr 2005 | 44% (4) | 11% (1) | 0% (0) | 22% (2) | 0% (0) | 11% (1) | 11% (1) | 0% (0) | 100% (9) |
| Apr 2006 | 35% (6) | 6% (1) | 0% (0) | 12% (2) | 0% (0) | 29% (5) | 18% (3) | 0% (0) | 100% (17) |
| Apr 2007 | 57% (8) | 14% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (2) | 14% (2) | 100% (14) |
| Apr 2008 | 65% (11) | 24% (4) | 0% (0) | 0% (0) | 6% (1) | 0% (0) | 6% (1) | 0% (0) | 100% (17) |
| Apr 2009 | 60% (6) | 0% (0) | 0% (0) | 0% (0) | 10% (1) | 0% (0) | 20% (2) | 10% (1) | 100% (10) |
| Apr 2010 | 36% (5) | 14% (2) | 0% (0) | 7% (1) | 7% (1) | 21% (3) | 14% (2) | 0% (0) | 100% (14) |
| Apr 2011 | 17% (2) | 0% (0) | 8% (1) | 0% (0) | 8% (1) | 17% (2) | 50% (6) | 0% (0) | 100% (12) |
| Apr 2012 | 27% (3) | 0% (0) | 0% (0) | 18% (2) | 0% (0) | 27% (3) | 27% (3) | 0% (0) | 100% (11) |
| Apr 2013 | 43% (6) | 0% (0) | 14% (2) | 7% (1) | 0% (0) | 14% (2) | 21% (3) | 0% (0) | 100% (14) |
| Apr 2014 | 0% (0) | 8% (1) | 31% (4) | 8% (1) | 0% (0) | 15% (2) | 15% (2) | 23% (3) | 100% (13) |
| Apr 2015 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 57% (4) | 29% (2) | 100% (7) |
| Apr 2016 | 13% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 13% (1) | 50% (4) | 25% (2) | 100% (8) |
| Apr 2017 | 0% (0) | 0% (0) | 17% (1) | 0% (0) | 0% (0) | 0% (0) | 50% (3) | 33% (2) | 100% (6) |
| Apr 2018 | 10% (1) | 0% (0) | 0% (0) | 10% (1) | 10% (1) | 20% (2) | 50% (5) | 0% (0) | 100% (10) |
| Apr 2019 | 11% (1) | 22% (2) | 0% (0) | 22% (2) | 0% (0) | 11% (1) | 22% (2) | 11% (1) | 100% (9) |
| Apr 2020 | 0% (0) | 10% (1) | 20% (2) | 10% (1) | 0% (0) | 10% (1) | 50% (5) | 0% (0) | 100% (10) |
| Apr 2021 | 31% (4) | 0% (0) | 8% (1) | 0% (0) | 0% (0) | 0% (0) | 62% (8) | 0% (0) | 100% (13) |
| Apr 2022 | 38% (5) | 23% (3) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 31% (4) | 8% (1) | 100% (13) |
| Apr 2023 | 20% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 20% (2) | 50% (5) | 10% (1) | 100% (10) |
| Apr 2024 | 22% (2) | 22% (2) | 0% (0) | 0% (0) | 11% (1) | 0% (0) | 33% (3) | 11% (1) | 100% (9) |
| Apr 2025 | 13% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 75% (6) | 13% (1) | 100% (8) |
| Apr 2026 | 25% (2) | 13% (1) | 0% (0) | 0% (0) | 0% (0) | 13% (1) | 50% (4) | 0% (0) | 100% (8) |
| Total | 33% (95) | 7% (21) | 4% (12) | 5% (15) | 2% (7) | 9% (28) | 33% (84) | 7% (17) | 100% (279) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
May
Table of remains recovered
| May_data$year_month | n | percent |
|---|---|---|
| May 2000 | 11 | 3.01% |
| May 2001 | 20 | 5.48% |
| May 2002 | 10 | 2.74% |
| May 2003 | 22 | 6.03% |
| May 2004 | 14 | 3.84% |
| May 2005 | 27 | 7.40% |
| May 2006 | 16 | 4.38% |
| May 2007 | 27 | 7.40% |
| May 2008 | 12 | 3.29% |
| May 2009 | 15 | 4.11% |
| May 2010 | 15 | 4.11% |
| May 2011 | 5 | 1.37% |
| May 2012 | 17 | 4.66% |
| May 2013 | 25 | 6.85% |
| May 2014 | 12 | 3.29% |
| May 2015 | 5 | 1.37% |
| May 2016 | 9 | 2.47% |
| May 2017 | 12 | 3.29% |
| May 2018 | 11 | 3.01% |
| May 2019 | 14 | 3.84% |
| May 2020 | 11 | 3.01% |
| May 2021 | 21 | 5.75% |
| May 2022 | 13 | 3.56% |
| May 2023 | 9 | 2.47% |
| May 2024 | 3 | 0.82% |
| May 2025 | 4 | 1.10% |
| May 2026 | 5 | 1.37% |
| Total | 365 | - |
Visualizing migrant deaths by May and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
May_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 122 | 33% | 35% |
| < 1 week | 49 | 13% | 14% |
| < 3 months | 23 | 6% | 7% |
| < 3 weeks | 24 | 7% | 7% |
| < 5 weeks | 30 | 8% | 9% |
| < 6-8 months | 27 | 7% | 8% |
| > 6-8 months | 75 | 21% | 21% |
| NA | 15 | 4% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| May 2000 | 64% (7) | 27% (3) | 9% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (11) |
| May 2001 | 50% (10) | 15% (3) | 0% (0) | 0% (0) | 35% (7) | 0% (0) | 0% (0) | 0% (0) | 100% (20) |
| May 2002 | 40% (4) | 50% (5) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 10% (1) | 0% (0) | 100% (10) |
| May 2003 | 45% (10) | 23% (5) | 0% (0) | 9% (2) | 5% (1) | 0% (0) | 18% (4) | 0% (0) | 100% (22) |
| May 2004 | 57% (8) | 7% (1) | 0% (0) | 14% (2) | 7% (1) | 0% (0) | 7% (1) | 7% (1) | 100% (14) |
| May 2005 | 37% (10) | 22% (6) | 4% (1) | 11% (3) | 22% (6) | 0% (0) | 0% (0) | 4% (1) | 100% (27) |
| May 2006 | 56% (9) | 6% (1) | 0% (0) | 6% (1) | 0% (0) | 13% (2) | 19% (3) | 0% (0) | 100% (16) |
| May 2007 | 33% (9) | 15% (4) | 7% (2) | 4% (1) | 7% (2) | 0% (0) | 30% (8) | 4% (1) | 100% (27) |
| May 2008 | 33% (4) | 0% (0) | 8% (1) | 17% (2) | 0% (0) | 25% (3) | 8% (1) | 8% (1) | 100% (12) |
| May 2009 | 53% (8) | 13% (2) | 13% (2) | 13% (2) | 7% (1) | 0% (0) | 0% (0) | 0% (0) | 100% (15) |
| May 2010 | 60% (9) | 7% (1) | 0% (0) | 0% (0) | 7% (1) | 13% (2) | 13% (2) | 0% (0) | 100% (15) |
| May 2011 | 20% (1) | 20% (1) | 20% (1) | 0% (0) | 0% (0) | 0% (0) | 40% (2) | 0% (0) | 100% (5) |
| May 2012 | 35% (6) | 6% (1) | 6% (1) | 18% (3) | 12% (2) | 6% (1) | 12% (2) | 6% (1) | 100% (17) |
| May 2013 | 28% (7) | 8% (2) | 4% (1) | 8% (2) | 8% (2) | 28% (7) | 8% (2) | 8% (2) | 100% (25) |
| May 2014 | 8% (1) | 0% (0) | 17% (2) | 25% (3) | 0% (0) | 8% (1) | 33% (4) | 8% (1) | 100% (12) |
| May 2015 | 0% (0) | 20% (1) | 0% (0) | 0% (0) | 0% (0) | 60% (3) | 20% (1) | 0% (0) | 100% (5) |
| May 2016 | 11% (1) | 0% (0) | 22% (2) | 0% (0) | 0% (0) | 22% (2) | 44% (4) | 0% (0) | 100% (9) |
| May 2017 | 0% (0) | 8% (1) | 8% (1) | 0% (0) | 17% (2) | 8% (1) | 50% (6) | 8% (1) | 100% (12) |
| May 2018 | 18% (2) | 18% (2) | 9% (1) | 0% (0) | 9% (1) | 18% (2) | 9% (1) | 18% (2) | 100% (11) |
| May 2019 | 7% (1) | 0% (0) | 7% (1) | 0% (0) | 14% (2) | 14% (2) | 50% (7) | 7% (1) | 100% (14) |
| May 2020 | 0% (0) | 18% (2) | 0% (0) | 9% (1) | 0% (0) | 0% (0) | 73% (8) | 0% (0) | 100% (11) |
| May 2021 | 24% (5) | 19% (4) | 0% (0) | 0% (0) | 5% (1) | 5% (1) | 33% (7) | 14% (3) | 100% (21) |
| May 2022 | 31% (4) | 8% (1) | 15% (2) | 8% (1) | 0% (0) | 0% (0) | 38% (5) | 0% (0) | 100% (13) |
| May 2023 | 22% (2) | 0% (0) | 22% (2) | 0% (0) | 11% (1) | 0% (0) | 44% (4) | 0% (0) | 100% (9) |
| May 2024 | 33% (1) | 33% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 33% (1) | 0% (0) | 100% (3) |
| May 2025 | 25% (1) | 0% (0) | 50% (2) | 25% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (4) |
| May 2026 | 40% (2) | 40% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 0% (0) | 100% (5) |
| Total | 31% (122) | 14% (49) | 8% (23) | 6% (24) | 6% (30) | 8% (27) | 23% (75) | 3% (15) | 100% (365) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
June
Table of remains recovered
| June_data$year_month | n | percent |
|---|---|---|
| Jun 2000 | 15 | 2.41% |
| Jun 2001 | 16 | 2.57% |
| Jun 2002 | 36 | 5.78% |
| Jun 2003 | 17 | 2.73% |
| Jun 2004 | 29 | 4.65% |
| Jun 2005 | 18 | 2.89% |
| Jun 2006 | 30 | 4.82% |
| Jun 2007 | 35 | 5.62% |
| Jun 2008 | 37 | 5.94% |
| Jun 2009 | 26 | 4.17% |
| Jun 2010 | 23 | 3.69% |
| Jun 2011 | 25 | 4.01% |
| Jun 2012 | 20 | 3.21% |
| Jun 2013 | 34 | 5.46% |
| Jun 2014 | 21 | 3.37% |
| Jun 2015 | 21 | 3.37% |
| Jun 2016 | 27 | 4.33% |
| Jun 2017 | 20 | 3.21% |
| Jun 2018 | 23 | 3.69% |
| Jun 2019 | 13 | 2.09% |
| Jun 2020 | 14 | 2.25% |
| Jun 2021 | 44 | 7.06% |
| Jun 2022 | 27 | 4.33% |
| Jun 2023 | 14 | 2.25% |
| Jun 2024 | 14 | 2.25% |
| Jun 2025 | 11 | 1.77% |
| Jun 2026 | 13 | 2.09% |
| Total | 623 | - |
Visualizing migrant deaths by June and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
June_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 233 | 37% | 39% |
| < 1 week | 118 | 19% | 20% |
| < 3 months | 24 | 4% | 4% |
| < 3 weeks | 61 | 10% | 10% |
| < 5 weeks | 62 | 10% | 10% |
| < 6-8 months | 36 | 6% | 6% |
| > 6-8 months | 69 | 11% | 11% |
| NA | 20 | 3% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Jun 2000 | 47% (7) | 47% (7) | 0% (0) | 7% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (15) |
| Jun 2001 | 44% (7) | 25% (4) | 0% (0) | 0% (0) | 25% (4) | 0% (0) | 6% (1) | 0% (0) | 100% (16) |
| Jun 2002 | 28% (10) | 28% (10) | 0% (0) | 22% (8) | 19% (7) | 0% (0) | 3% (1) | 0% (0) | 100% (36) |
| Jun 2003 | 41% (7) | 29% (5) | 0% (0) | 0% (0) | 12% (2) | 0% (0) | 18% (3) | 0% (0) | 100% (17) |
| Jun 2004 | 76% (22) | 10% (3) | 0% (0) | 3% (1) | 3% (1) | 0% (0) | 3% (1) | 3% (1) | 100% (29) |
| Jun 2005 | 22% (4) | 11% (2) | 0% (0) | 39% (7) | 17% (3) | 0% (0) | 6% (1) | 6% (1) | 100% (18) |
| Jun 2006 | 53% (16) | 10% (3) | 3% (1) | 13% (4) | 13% (4) | 0% (0) | 7% (2) | 0% (0) | 100% (30) |
| Jun 2007 | 57% (20) | 23% (8) | 0% (0) | 9% (3) | 9% (3) | 0% (0) | 3% (1) | 0% (0) | 100% (35) |
| Jun 2008 | 41% (15) | 32% (12) | 0% (0) | 11% (4) | 0% (0) | 11% (4) | 5% (2) | 0% (0) | 100% (37) |
| Jun 2009 | 69% (18) | 4% (1) | 0% (0) | 8% (2) | 4% (1) | 0% (0) | 12% (3) | 4% (1) | 100% (26) |
| Jun 2010 | 57% (13) | 22% (5) | 4% (1) | 0% (0) | 4% (1) | 0% (0) | 9% (2) | 4% (1) | 100% (23) |
| Jun 2011 | 44% (11) | 20% (5) | 4% (1) | 4% (1) | 4% (1) | 8% (2) | 16% (4) | 0% (0) | 100% (25) |
| Jun 2012 | 40% (8) | 20% (4) | 0% (0) | 0% (0) | 30% (6) | 5% (1) | 5% (1) | 0% (0) | 100% (20) |
| Jun 2013 | 18% (6) | 12% (4) | 12% (4) | 9% (3) | 12% (4) | 15% (5) | 9% (3) | 15% (5) | 100% (34) |
| Jun 2014 | 14% (3) | 10% (2) | 10% (2) | 10% (2) | 19% (4) | 24% (5) | 5% (1) | 10% (2) | 100% (21) |
| Jun 2015 | 29% (6) | 10% (2) | 5% (1) | 19% (4) | 0% (0) | 5% (1) | 24% (5) | 10% (2) | 100% (21) |
| Jun 2016 | 26% (7) | 22% (6) | 4% (1) | 22% (6) | 11% (3) | 0% (0) | 4% (1) | 11% (3) | 100% (27) |
| Jun 2017 | 10% (2) | 15% (3) | 10% (2) | 10% (2) | 5% (1) | 40% (8) | 10% (2) | 0% (0) | 100% (20) |
| Jun 2018 | 9% (2) | 17% (4) | 0% (0) | 17% (4) | 9% (2) | 17% (4) | 30% (7) | 0% (0) | 100% (23) |
| Jun 2019 | 23% (3) | 8% (1) | 23% (3) | 8% (1) | 0% (0) | 8% (1) | 23% (3) | 8% (1) | 100% (13) |
| Jun 2020 | 21% (3) | 7% (1) | 0% (0) | 0% (0) | 21% (3) | 7% (1) | 29% (4) | 14% (2) | 100% (14) |
| Jun 2021 | 34% (15) | 32% (14) | 7% (3) | 5% (2) | 11% (5) | 2% (1) | 9% (4) | 0% (0) | 100% (44) |
| Jun 2022 | 30% (8) | 15% (4) | 15% (4) | 11% (3) | 7% (2) | 0% (0) | 22% (6) | 0% (0) | 100% (27) |
| Jun 2023 | 36% (5) | 7% (1) | 0% (0) | 7% (1) | 14% (2) | 14% (2) | 14% (2) | 7% (1) | 100% (14) |
| Jun 2024 | 57% (8) | 14% (2) | 0% (0) | 0% (0) | 21% (3) | 0% (0) | 7% (1) | 0% (0) | 100% (14) |
| Jun 2025 | 18% (2) | 0% (0) | 9% (1) | 9% (1) | 0% (0) | 9% (1) | 55% (6) | 0% (0) | 100% (11) |
| Jun 2026 | 38% (5) | 38% (5) | 0% (0) | 8% (1) | 0% (0) | 0% (0) | 15% (2) | 0% (0) | 100% (13) |
| Total | 36% (233) | 18% (118) | 4% (24) | 9% (61) | 10% (62) | 6% (36) | 13% (69) | 3% (20) | 100% (623) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
July
Table of remains recovered
| July_data$year_month | n | percent |
|---|---|---|
| Jul 2000 | 5 | 0.67% |
| Jul 2001 | 16 | 2.14% |
| Jul 2002 | 23 | 3.08% |
| Jul 2003 | 39 | 5.22% |
| Jul 2004 | 30 | 4.02% |
| Jul 2005 | 69 | 9.24% |
| Jul 2006 | 35 | 4.69% |
| Jul 2007 | 46 | 6.16% |
| Jul 2008 | 21 | 2.81% |
| Jul 2009 | 34 | 4.55% |
| Jul 2010 | 57 | 7.63% |
| Jul 2011 | 21 | 2.81% |
| Jul 2012 | 30 | 4.02% |
| Jul 2013 | 35 | 4.69% |
| Jul 2014 | 14 | 1.87% |
| Jul 2015 | 16 | 2.14% |
| Jul 2016 | 26 | 3.48% |
| Jul 2017 | 14 | 1.87% |
| Jul 2018 | 13 | 1.74% |
| Jul 2019 | 19 | 2.54% |
| Jul 2020 | 25 | 3.35% |
| Jul 2021 | 21 | 2.81% |
| Jul 2022 | 31 | 4.15% |
| Jul 2023 | 42 | 5.62% |
| Jul 2024 | 36 | 4.82% |
| Jul 2025 | 10 | 1.34% |
| Jul 2026 | 19 | 2.54% |
| Total | 747 | - |
Visualizing migrant deaths by July and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
July_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 235 | 31% | 33% |
| < 1 week | 121 | 16% | 17% |
| < 3 months | 41 | 5% | 6% |
| < 3 weeks | 84 | 11% | 12% |
| < 5 weeks | 118 | 16% | 17% |
| < 6-8 months | 26 | 3% | 4% |
| > 6-8 months | 90 | 12% | 13% |
| NA | 32 | 4% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Jul 2000 | 20% (1) | 40% (2) | 20% (1) | 0% (0) | 20% (1) | 0% (0) | 0% (0) | 0% (0) | 100% (5) |
| Jul 2001 | 50% (8) | 31% (5) | 0% (0) | 0% (0) | 13% (2) | 6% (1) | 0% (0) | 0% (0) | 100% (16) |
| Jul 2002 | 43% (10) | 17% (4) | 9% (2) | 9% (2) | 17% (4) | 0% (0) | 0% (0) | 4% (1) | 100% (23) |
| Jul 2003 | 31% (12) | 18% (7) | 0% (0) | 18% (7) | 18% (7) | 0% (0) | 8% (3) | 8% (3) | 100% (39) |
| Jul 2004 | 37% (11) | 23% (7) | 3% (1) | 3% (1) | 7% (2) | 3% (1) | 7% (2) | 17% (5) | 100% (30) |
| Jul 2005 | 45% (31) | 14% (10) | 3% (2) | 22% (15) | 16% (11) | 0% (0) | 0% (0) | 0% (0) | 100% (69) |
| Jul 2006 | 40% (14) | 29% (10) | 3% (1) | 9% (3) | 6% (2) | 6% (2) | 3% (1) | 6% (2) | 100% (35) |
| Jul 2007 | 30% (14) | 11% (5) | 15% (7) | 13% (6) | 24% (11) | 0% (0) | 7% (3) | 0% (0) | 100% (46) |
| Jul 2008 | 29% (6) | 33% (7) | 0% (0) | 0% (0) | 14% (3) | 10% (2) | 10% (2) | 5% (1) | 100% (21) |
| Jul 2009 | 50% (17) | 12% (4) | 3% (1) | 12% (4) | 9% (3) | 3% (1) | 9% (3) | 3% (1) | 100% (34) |
| Jul 2010 | 33% (19) | 21% (12) | 5% (3) | 16% (9) | 16% (9) | 4% (2) | 5% (3) | 0% (0) | 100% (57) |
| Jul 2011 | 24% (5) | 19% (4) | 0% (0) | 0% (0) | 19% (4) | 5% (1) | 29% (6) | 5% (1) | 100% (21) |
| Jul 2012 | 17% (5) | 23% (7) | 7% (2) | 13% (4) | 17% (5) | 0% (0) | 17% (5) | 7% (2) | 100% (30) |
| Jul 2013 | 17% (6) | 6% (2) | 9% (3) | 20% (7) | 17% (6) | 9% (3) | 11% (4) | 11% (4) | 100% (35) |
| Jul 2014 | 14% (2) | 0% (0) | 0% (0) | 21% (3) | 14% (2) | 21% (3) | 21% (3) | 7% (1) | 100% (14) |
| Jul 2015 | 6% (1) | 6% (1) | 0% (0) | 31% (5) | 13% (2) | 0% (0) | 31% (5) | 13% (2) | 100% (16) |
| Jul 2016 | 35% (9) | 12% (3) | 8% (2) | 8% (2) | 19% (5) | 8% (2) | 0% (0) | 12% (3) | 100% (26) |
| Jul 2017 | 7% (1) | 0% (0) | 7% (1) | 0% (0) | 36% (5) | 7% (1) | 29% (4) | 14% (2) | 100% (14) |
| Jul 2018 | 31% (4) | 0% (0) | 0% (0) | 0% (0) | 38% (5) | 8% (1) | 23% (3) | 0% (0) | 100% (13) |
| Jul 2019 | 16% (3) | 26% (5) | 5% (1) | 16% (3) | 16% (3) | 0% (0) | 21% (4) | 0% (0) | 100% (19) |
| Jul 2020 | 36% (9) | 16% (4) | 4% (1) | 0% (0) | 24% (6) | 8% (2) | 12% (3) | 0% (0) | 100% (25) |
| Jul 2021 | 14% (3) | 10% (2) | 5% (1) | 24% (5) | 24% (5) | 5% (1) | 10% (2) | 10% (2) | 100% (21) |
| Jul 2022 | 19% (6) | 16% (5) | 6% (2) | 10% (3) | 13% (4) | 3% (1) | 29% (9) | 3% (1) | 100% (31) |
| Jul 2023 | 50% (21) | 10% (4) | 10% (4) | 5% (2) | 10% (4) | 5% (2) | 12% (5) | 0% (0) | 100% (42) |
| Jul 2024 | 28% (10) | 22% (8) | 8% (3) | 8% (3) | 8% (3) | 0% (0) | 22% (8) | 3% (1) | 100% (36) |
| Jul 2025 | 20% (2) | 10% (1) | 20% (2) | 0% (0) | 10% (1) | 0% (0) | 40% (4) | 0% (0) | 100% (10) |
| Jul 2026 | 26% (5) | 11% (2) | 5% (1) | 0% (0) | 16% (3) | 0% (0) | 42% (8) | 0% (0) | 100% (19) |
| Total | 28% (235) | 16% (121) | 6% (41) | 10% (84) | 17% (118) | 4% (26) | 15% (90) | 5% (32) | 100% (747) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
August
Table of remains recovered
| August_data$year_month | n | percent |
|---|---|---|
| Aug 2000 | 7 | 1.55% |
| Aug 2001 | 4 | 0.88% |
| Aug 2002 | 29 | 6.42% |
| Aug 2003 | 22 | 4.87% |
| Aug 2004 | 21 | 4.65% |
| Aug 2005 | 19 | 4.20% |
| Aug 2006 | 13 | 2.88% |
| Aug 2007 | 22 | 4.87% |
| Aug 2008 | 20 | 4.42% |
| Aug 2009 | 26 | 5.75% |
| Aug 2010 | 21 | 4.65% |
| Aug 2011 | 26 | 5.75% |
| Aug 2012 | 12 | 2.65% |
| Aug 2013 | 16 | 3.54% |
| Aug 2014 | 9 | 1.99% |
| Aug 2015 | 30 | 6.64% |
| Aug 2016 | 17 | 3.76% |
| Aug 2017 | 10 | 2.21% |
| Aug 2018 | 14 | 3.10% |
| Aug 2019 | 13 | 2.88% |
| Aug 2020 | 24 | 5.31% |
| Aug 2021 | 16 | 3.54% |
| Aug 2022 | 10 | 2.21% |
| Aug 2023 | 25 | 5.53% |
| Aug 2024 | 15 | 3.32% |
| Aug 2025 | 11 | 2.43% |
| Total | 452 | - |
Visualizing migrant deaths by August and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
August_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 137 | 30% | 32% |
| < 1 week | 63 | 14% | 15% |
| < 3 months | 45 | 10% | 10% |
| < 3 weeks | 37 | 8% | 9% |
| < 5 weeks | 68 | 15% | 16% |
| < 6-8 months | 29 | 6% | 7% |
| > 6-8 months | 55 | 12% | 13% |
| NA | 18 | 4% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Aug 2000 | 86% (6) | 14% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (7) |
| Aug 2001 | 75% (3) | 25% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (4) |
| Aug 2002 | 48% (14) | 17% (5) | 3% (1) | 3% (1) | 21% (6) | 0% (0) | 7% (2) | 0% (0) | 100% (29) |
| Aug 2003 | 36% (8) | 9% (2) | 9% (2) | 9% (2) | 23% (5) | 5% (1) | 9% (2) | 0% (0) | 100% (22) |
| Aug 2004 | 38% (8) | 0% (0) | 5% (1) | 10% (2) | 10% (2) | 5% (1) | 5% (1) | 29% (6) | 100% (21) |
| Aug 2005 | 11% (2) | 11% (2) | 16% (3) | 0% (0) | 21% (4) | 26% (5) | 16% (3) | 0% (0) | 100% (19) |
| Aug 2006 | 23% (3) | 23% (3) | 8% (1) | 0% (0) | 8% (1) | 8% (1) | 15% (2) | 15% (2) | 100% (13) |
| Aug 2007 | 50% (11) | 9% (2) | 5% (1) | 14% (3) | 9% (2) | 5% (1) | 9% (2) | 0% (0) | 100% (22) |
| Aug 2008 | 50% (10) | 20% (4) | 0% (0) | 15% (3) | 5% (1) | 10% (2) | 0% (0) | 0% (0) | 100% (20) |
| Aug 2009 | 38% (10) | 23% (6) | 8% (2) | 0% (0) | 19% (5) | 0% (0) | 8% (2) | 4% (1) | 100% (26) |
| Aug 2010 | 14% (3) | 14% (3) | 29% (6) | 5% (1) | 14% (3) | 14% (3) | 10% (2) | 0% (0) | 100% (21) |
| Aug 2011 | 19% (5) | 8% (2) | 19% (5) | 19% (5) | 8% (2) | 12% (3) | 15% (4) | 0% (0) | 100% (26) |
| Aug 2012 | 8% (1) | 8% (1) | 17% (2) | 25% (3) | 17% (2) | 17% (2) | 8% (1) | 0% (0) | 100% (12) |
| Aug 2013 | 19% (3) | 6% (1) | 19% (3) | 0% (0) | 13% (2) | 0% (0) | 19% (3) | 25% (4) | 100% (16) |
| Aug 2014 | 11% (1) | 0% (0) | 11% (1) | 11% (1) | 22% (2) | 22% (2) | 22% (2) | 0% (0) | 100% (9) |
| Aug 2015 | 17% (5) | 20% (6) | 7% (2) | 10% (3) | 27% (8) | 3% (1) | 10% (3) | 7% (2) | 100% (30) |
| Aug 2016 | 12% (2) | 18% (3) | 18% (3) | 0% (0) | 18% (3) | 6% (1) | 18% (3) | 12% (2) | 100% (17) |
| Aug 2017 | 30% (3) | 10% (1) | 0% (0) | 10% (1) | 10% (1) | 0% (0) | 40% (4) | 0% (0) | 100% (10) |
| Aug 2018 | 29% (4) | 0% (0) | 29% (4) | 14% (2) | 14% (2) | 7% (1) | 0% (0) | 7% (1) | 100% (14) |
| Aug 2019 | 8% (1) | 15% (2) | 8% (1) | 38% (5) | 15% (2) | 0% (0) | 15% (2) | 0% (0) | 100% (13) |
| Aug 2020 | 33% (8) | 25% (6) | 0% (0) | 4% (1) | 21% (5) | 0% (0) | 17% (4) | 0% (0) | 100% (24) |
| Aug 2021 | 31% (5) | 6% (1) | 6% (1) | 13% (2) | 19% (3) | 6% (1) | 19% (3) | 0% (0) | 100% (16) |
| Aug 2022 | 20% (2) | 30% (3) | 20% (2) | 0% (0) | 10% (1) | 10% (1) | 10% (1) | 0% (0) | 100% (10) |
| Aug 2023 | 32% (8) | 20% (5) | 12% (3) | 8% (2) | 20% (5) | 4% (1) | 4% (1) | 0% (0) | 100% (25) |
| Aug 2024 | 60% (9) | 13% (2) | 7% (1) | 0% (0) | 0% (0) | 13% (2) | 7% (1) | 0% (0) | 100% (15) |
| Aug 2025 | 18% (2) | 9% (1) | 0% (0) | 0% (0) | 9% (1) | 0% (0) | 64% (7) | 0% (0) | 100% (11) |
| Total | 31% (137) | 14% (63) | 10% (45) | 8% (37) | 14% (68) | 7% (29) | 13% (55) | 4% (18) | 100% (452) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
September
Table of remains recovered
| September_data$year_month | n | percent |
|---|---|---|
| Sep 2000 | 7 | 1.82% |
| Sep 2001 | 7 | 1.82% |
| Sep 2002 | 21 | 5.45% |
| Sep 2003 | 12 | 3.12% |
| Sep 2004 | 19 | 4.94% |
| Sep 2005 | 19 | 4.94% |
| Sep 2006 | 12 | 3.12% |
| Sep 2007 | 6 | 1.56% |
| Sep 2008 | 10 | 2.60% |
| Sep 2009 | 15 | 3.90% |
| Sep 2010 | 14 | 3.64% |
| Sep 2011 | 14 | 3.64% |
| Sep 2012 | 6 | 1.56% |
| Sep 2013 | 11 | 2.86% |
| Sep 2014 | 11 | 2.86% |
| Sep 2015 | 20 | 5.19% |
| Sep 2016 | 22 | 5.71% |
| Sep 2017 | 13 | 3.38% |
| Sep 2018 | 12 | 3.12% |
| Sep 2019 | 14 | 3.64% |
| Sep 2020 | 35 | 9.09% |
| Sep 2021 | 26 | 6.75% |
| Sep 2022 | 13 | 3.38% |
| Sep 2023 | 27 | 7.01% |
| Sep 2024 | 17 | 4.42% |
| Sep 2025 | 2 | 0.52% |
| Total | 385 | - |
Visualizing migrant deaths by September and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
September_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 86 | 22% | 23% |
| < 1 week | 43 | 11% | 12% |
| < 3 months | 45 | 12% | 12% |
| < 3 weeks | 28 | 7% | 8% |
| < 5 weeks | 50 | 13% | 14% |
| < 6-8 months | 34 | 9% | 9% |
| > 6-8 months | 84 | 22% | 23% |
| NA | 15 | 4% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Sep 2000 | 57% (4) | 14% (1) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 0% (0) | 14% (1) | 100% (7) |
| Sep 2001 | 57% (4) | 29% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 0% (0) | 100% (7) |
| Sep 2002 | 33% (7) | 24% (5) | 5% (1) | 10% (2) | 5% (1) | 0% (0) | 19% (4) | 5% (1) | 100% (21) |
| Sep 2003 | 58% (7) | 17% (2) | 8% (1) | 0% (0) | 8% (1) | 8% (1) | 0% (0) | 0% (0) | 100% (12) |
| Sep 2004 | 58% (11) | 11% (2) | 5% (1) | 0% (0) | 11% (2) | 5% (1) | 11% (2) | 0% (0) | 100% (19) |
| Sep 2005 | 42% (8) | 0% (0) | 0% (0) | 16% (3) | 11% (2) | 11% (2) | 11% (2) | 11% (2) | 100% (19) |
| Sep 2006 | 25% (3) | 8% (1) | 8% (1) | 0% (0) | 0% (0) | 25% (3) | 33% (4) | 0% (0) | 100% (12) |
| Sep 2007 | 50% (3) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 17% (1) | 33% (2) | 0% (0) | 100% (6) |
| Sep 2008 | 0% (0) | 20% (2) | 20% (2) | 10% (1) | 10% (1) | 10% (1) | 30% (3) | 0% (0) | 100% (10) |
| Sep 2009 | 20% (3) | 7% (1) | 20% (3) | 0% (0) | 20% (3) | 7% (1) | 27% (4) | 0% (0) | 100% (15) |
| Sep 2010 | 29% (4) | 14% (2) | 21% (3) | 7% (1) | 0% (0) | 7% (1) | 21% (3) | 0% (0) | 100% (14) |
| Sep 2011 | 14% (2) | 0% (0) | 14% (2) | 14% (2) | 36% (5) | 0% (0) | 21% (3) | 0% (0) | 100% (14) |
| Sep 2012 | 0% (0) | 0% (0) | 17% (1) | 33% (2) | 0% (0) | 50% (3) | 0% (0) | 0% (0) | 100% (6) |
| Sep 2013 | 9% (1) | 9% (1) | 18% (2) | 0% (0) | 9% (1) | 18% (2) | 18% (2) | 18% (2) | 100% (11) |
| Sep 2014 | 18% (2) | 9% (1) | 9% (1) | 9% (1) | 18% (2) | 9% (1) | 18% (2) | 9% (1) | 100% (11) |
| Sep 2015 | 0% (0) | 0% (0) | 15% (3) | 10% (2) | 45% (9) | 10% (2) | 15% (3) | 5% (1) | 100% (20) |
| Sep 2016 | 5% (1) | 0% (0) | 23% (5) | 5% (1) | 18% (4) | 14% (3) | 27% (6) | 9% (2) | 100% (22) |
| Sep 2017 | 15% (2) | 0% (0) | 8% (1) | 0% (0) | 23% (3) | 23% (3) | 15% (2) | 15% (2) | 100% (13) |
| Sep 2018 | 25% (3) | 8% (1) | 8% (1) | 17% (2) | 17% (2) | 25% (3) | 0% (0) | 0% (0) | 100% (12) |
| Sep 2019 | 0% (0) | 14% (2) | 14% (2) | 29% (4) | 14% (2) | 0% (0) | 29% (4) | 0% (0) | 100% (14) |
| Sep 2020 | 20% (7) | 9% (3) | 20% (7) | 3% (1) | 14% (5) | 6% (2) | 29% (10) | 0% (0) | 100% (35) |
| Sep 2021 | 23% (6) | 19% (5) | 8% (2) | 8% (2) | 0% (0) | 4% (1) | 35% (9) | 4% (1) | 100% (26) |
| Sep 2022 | 0% (0) | 23% (3) | 8% (1) | 8% (1) | 23% (3) | 0% (0) | 38% (5) | 0% (0) | 100% (13) |
| Sep 2023 | 22% (6) | 19% (5) | 15% (4) | 7% (2) | 11% (3) | 4% (1) | 22% (6) | 0% (0) | 100% (27) |
| Sep 2024 | 6% (1) | 24% (4) | 6% (1) | 6% (1) | 6% (1) | 6% (1) | 35% (6) | 12% (2) | 100% (17) |
| Sep 2025 | 50% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (1) | 0% (0) | 100% (2) |
| Total | 25% (86) | 11% (43) | 10% (45) | 7% (28) | 11% (50) | 10% (34) | 21% (84) | 4% (15) | 100% (385) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
October
Table of remains recovered
| October_data$year_month | n | percent |
|---|---|---|
| Oct 2000 | 4 | 1.63% |
| Oct 2001 | 1 | 0.41% |
| Oct 2002 | 8 | 3.27% |
| Oct 2003 | 8 | 3.27% |
| Oct 2004 | 10 | 4.08% |
| Oct 2005 | 5 | 2.04% |
| Oct 2006 | 7 | 2.86% |
| Oct 2007 | 12 | 4.90% |
| Oct 2008 | 9 | 3.67% |
| Oct 2009 | 8 | 3.27% |
| Oct 2010 | 16 | 6.53% |
| Oct 2011 | 15 | 6.12% |
| Oct 2012 | 8 | 3.27% |
| Oct 2013 | 4 | 1.63% |
| Oct 2014 | 13 | 5.31% |
| Oct 2015 | 10 | 4.08% |
| Oct 2016 | 14 | 5.71% |
| Oct 2017 | 6 | 2.45% |
| Oct 2018 | 10 | 4.08% |
| Oct 2019 | 14 | 5.71% |
| Oct 2020 | 17 | 6.94% |
| Oct 2021 | 12 | 4.90% |
| Oct 2022 | 6 | 2.45% |
| Oct 2023 | 14 | 5.71% |
| Oct 2024 | 9 | 3.67% |
| Oct 2025 | 5 | 2.04% |
| Total | 245 | - |
Visualizing migrant deaths by October and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
October_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 40 | 16% | 17% |
| < 1 week | 27 | 11% | 12% |
| < 3 months | 40 | 16% | 17% |
| < 3 weeks | 9 | 4% | 4% |
| < 5 weeks | 23 | 9% | 10% |
| < 6-8 months | 30 | 12% | 13% |
| > 6-8 months | 62 | 25% | 27% |
| NA | 14 | 6% | - |
One thing that may be of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Oct 2000 | 75% (3) | 25% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (4) |
| Oct 2001 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (1) | 0% (0) | 0% (0) | 100% (1) |
| Oct 2002 | 25% (2) | 0% (0) | 0% (0) | 25% (2) | 0% (0) | 0% (0) | 50% (4) | 0% (0) | 100% (8) |
| Oct 2003 | 0% (0) | 25% (2) | 0% (0) | 25% (2) | 0% (0) | 13% (1) | 13% (1) | 25% (2) | 100% (8) |
| Oct 2004 | 30% (3) | 10% (1) | 10% (1) | 0% (0) | 0% (0) | 20% (2) | 30% (3) | 0% (0) | 100% (10) |
| Oct 2005 | 20% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 40% (2) | 40% (2) | 100% (5) |
| Oct 2006 | 29% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 14% (1) | 57% (4) | 0% (0) | 100% (7) |
| Oct 2007 | 8% (1) | 8% (1) | 0% (0) | 0% (0) | 8% (1) | 50% (6) | 25% (3) | 0% (0) | 100% (12) |
| Oct 2008 | 56% (5) | 22% (2) | 0% (0) | 0% (0) | 11% (1) | 0% (0) | 11% (1) | 0% (0) | 100% (9) |
| Oct 2009 | 0% (0) | 0% (0) | 38% (3) | 0% (0) | 13% (1) | 13% (1) | 38% (3) | 0% (0) | 100% (8) |
| Oct 2010 | 25% (4) | 19% (3) | 25% (4) | 0% (0) | 6% (1) | 6% (1) | 19% (3) | 0% (0) | 100% (16) |
| Oct 2011 | 13% (2) | 7% (1) | 27% (4) | 0% (0) | 7% (1) | 13% (2) | 33% (5) | 0% (0) | 100% (15) |
| Oct 2012 | 0% (0) | 0% (0) | 38% (3) | 13% (1) | 13% (1) | 0% (0) | 38% (3) | 0% (0) | 100% (8) |
| Oct 2013 | 50% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (2) | 0% (0) | 0% (0) | 100% (4) |
| Oct 2014 | 8% (1) | 0% (0) | 15% (2) | 8% (1) | 8% (1) | 15% (2) | 23% (3) | 23% (3) | 100% (13) |
| Oct 2015 | 20% (2) | 10% (1) | 20% (2) | 10% (1) | 10% (1) | 0% (0) | 20% (2) | 10% (1) | 100% (10) |
| Oct 2016 | 0% (0) | 0% (0) | 43% (6) | 0% (0) | 7% (1) | 14% (2) | 29% (4) | 7% (1) | 100% (14) |
| Oct 2017 | 17% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (3) | 33% (2) | 0% (0) | 100% (6) |
| Oct 2018 | 10% (1) | 0% (0) | 20% (2) | 10% (1) | 10% (1) | 10% (1) | 30% (3) | 10% (1) | 100% (10) |
| Oct 2019 | 7% (1) | 14% (2) | 29% (4) | 0% (0) | 29% (4) | 0% (0) | 21% (3) | 0% (0) | 100% (14) |
| Oct 2020 | 24% (4) | 18% (3) | 24% (4) | 0% (0) | 6% (1) | 0% (0) | 18% (3) | 12% (2) | 100% (17) |
| Oct 2021 | 8% (1) | 33% (4) | 8% (1) | 8% (1) | 8% (1) | 8% (1) | 17% (2) | 8% (1) | 100% (12) |
| Oct 2022 | 17% (1) | 0% (0) | 0% (0) | 0% (0) | 33% (2) | 0% (0) | 50% (3) | 0% (0) | 100% (6) |
| Oct 2023 | 7% (1) | 36% (5) | 14% (2) | 0% (0) | 14% (2) | 21% (3) | 7% (1) | 0% (0) | 100% (14) |
| Oct 2024 | 22% (2) | 11% (1) | 0% (0) | 0% (0) | 33% (3) | 11% (1) | 11% (1) | 11% (1) | 100% (9) |
| Oct 2025 | 0% (0) | 0% (0) | 40% (2) | 0% (0) | 0% (0) | 0% (0) | 60% (3) | 0% (0) | 100% (5) |
| Total | 18% (40) | 9% (27) | 13% (40) | 4% (9) | 8% (23) | 16% (30) | 26% (62) | 6% (14) | 100% (245) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
November
Table of remains recovered
| November_data$year_month | n | percent |
|---|---|---|
| Nov 2000 | 3 | 1.20% |
| Nov 2001 | 8 | 3.19% |
| Nov 2002 | 4 | 1.59% |
| Nov 2003 | 12 | 4.78% |
| Nov 2004 | 9 | 3.59% |
| Nov 2005 | 7 | 2.79% |
| Nov 2006 | 7 | 2.79% |
| Nov 2007 | 9 | 3.59% |
| Nov 2008 | 5 | 1.99% |
| Nov 2009 | 15 | 5.98% |
| Nov 2010 | 13 | 5.18% |
| Nov 2011 | 17 | 6.77% |
| Nov 2012 | 12 | 4.78% |
| Nov 2013 | 8 | 3.19% |
| Nov 2014 | 8 | 3.19% |
| Nov 2015 | 10 | 3.98% |
| Nov 2016 | 11 | 4.38% |
| Nov 2017 | 8 | 3.19% |
| Nov 2018 | 3 | 1.20% |
| Nov 2019 | 6 | 2.39% |
| Nov 2020 | 14 | 5.58% |
| Nov 2021 | 16 | 6.37% |
| Nov 2022 | 11 | 4.38% |
| Nov 2023 | 14 | 5.58% |
| Nov 2024 | 10 | 3.98% |
| Nov 2025 | 11 | 4.38% |
| Total | 251 | - |
Visualizing migrant deaths by November and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
November_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 33 | 13% | 14% |
| < 1 week | 9 | 4% | 4% |
| < 3 months | 38 | 15% | 16% |
| < 3 weeks | 5 | 2% | 2% |
| < 5 weeks | 11 | 4% | 5% |
| < 6-8 months | 47 | 19% | 20% |
| > 6-8 months | 91 | 36% | 39% |
| NA | 17 | 7% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Nov 2000 | 67% (2) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 33% (1) | 0% (0) | 0% (0) | 100% (3) |
| Nov 2001 | 25% (2) | 0% (0) | 13% (1) | 0% (0) | 13% (1) | 13% (1) | 13% (1) | 25% (2) | 100% (8) |
| Nov 2002 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 25% (1) | 25% (1) | 50% (2) | 100% (4) |
| Nov 2003 | 42% (5) | 0% (0) | 17% (2) | 0% (0) | 0% (0) | 8% (1) | 17% (2) | 17% (2) | 100% (12) |
| Nov 2004 | 11% (1) | 0% (0) | 11% (1) | 0% (0) | 11% (1) | 33% (3) | 33% (3) | 0% (0) | 100% (9) |
| Nov 2005 | 43% (3) | 14% (1) | 0% (0) | 14% (1) | 0% (0) | 0% (0) | 29% (2) | 0% (0) | 100% (7) |
| Nov 2006 | 14% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 43% (3) | 43% (3) | 0% (0) | 100% (7) |
| Nov 2007 | 0% (0) | 0% (0) | 0% (0) | 11% (1) | 0% (0) | 44% (4) | 44% (4) | 0% (0) | 100% (9) |
| Nov 2008 | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 0% (0) | 20% (1) | 60% (3) | 0% (0) | 100% (5) |
| Nov 2009 | 27% (4) | 0% (0) | 7% (1) | 0% (0) | 20% (3) | 7% (1) | 33% (5) | 7% (1) | 100% (15) |
| Nov 2010 | 23% (3) | 0% (0) | 31% (4) | 0% (0) | 8% (1) | 38% (5) | 0% (0) | 0% (0) | 100% (13) |
| Nov 2011 | 18% (3) | 6% (1) | 29% (5) | 0% (0) | 0% (0) | 18% (3) | 29% (5) | 0% (0) | 100% (17) |
| Nov 2012 | 0% (0) | 8% (1) | 25% (3) | 0% (0) | 8% (1) | 42% (5) | 8% (1) | 8% (1) | 100% (12) |
| Nov 2013 | 13% (1) | 0% (0) | 13% (1) | 13% (1) | 13% (1) | 0% (0) | 25% (2) | 25% (2) | 100% (8) |
| Nov 2014 | 13% (1) | 13% (1) | 25% (2) | 0% (0) | 0% (0) | 13% (1) | 38% (3) | 0% (0) | 100% (8) |
| Nov 2015 | 0% (0) | 0% (0) | 10% (1) | 0% (0) | 20% (2) | 50% (5) | 10% (1) | 10% (1) | 100% (10) |
| Nov 2016 | 9% (1) | 0% (0) | 18% (2) | 0% (0) | 0% (0) | 9% (1) | 45% (5) | 18% (2) | 100% (11) |
| Nov 2017 | 13% (1) | 0% (0) | 13% (1) | 0% (0) | 0% (0) | 25% (2) | 38% (3) | 13% (1) | 100% (8) |
| Nov 2018 | 33% (1) | 0% (0) | 33% (1) | 0% (0) | 0% (0) | 0% (0) | 33% (1) | 0% (0) | 100% (3) |
| Nov 2019 | 0% (0) | 0% (0) | 67% (4) | 0% (0) | 0% (0) | 0% (0) | 33% (2) | 0% (0) | 100% (6) |
| Nov 2020 | 0% (0) | 7% (1) | 7% (1) | 7% (1) | 0% (0) | 0% (0) | 71% (10) | 7% (1) | 100% (14) |
| Nov 2021 | 19% (3) | 0% (0) | 19% (3) | 0% (0) | 0% (0) | 25% (4) | 31% (5) | 6% (1) | 100% (16) |
| Nov 2022 | 0% (0) | 9% (1) | 9% (1) | 0% (0) | 0% (0) | 9% (1) | 73% (8) | 0% (0) | 100% (11) |
| Nov 2023 | 7% (1) | 21% (3) | 7% (1) | 0% (0) | 0% (0) | 29% (4) | 36% (5) | 0% (0) | 100% (14) |
| Nov 2024 | 0% (0) | 0% (0) | 30% (3) | 0% (0) | 0% (0) | 0% (0) | 70% (7) | 0% (0) | 100% (10) |
| Nov 2025 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 9% (1) | 0% (0) | 82% (9) | 9% (1) | 100% (11) |
| Total | 14% (33) | 3% (9) | 15% (38) | 3% (5) | 4% (11) | 19% (47) | 35% (91) | 7% (17) | 100% (251) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
December
Table of remains recovered
| December_data$year_month | n | percent |
|---|---|---|
| Dec 2000 | 1 | 0.45% |
| Dec 2001 | 2 | 0.90% |
| Dec 2002 | 3 | 1.36% |
| Dec 2003 | 6 | 2.71% |
| Dec 2004 | 5 | 2.26% |
| Dec 2005 | 11 | 4.98% |
| Dec 2006 | 8 | 3.62% |
| Dec 2007 | 8 | 3.62% |
| Dec 2008 | 17 | 7.69% |
| Dec 2009 | 16 | 7.24% |
| Dec 2010 | 10 | 4.52% |
| Dec 2011 | 12 | 5.43% |
| Dec 2012 | 11 | 4.98% |
| Dec 2013 | 7 | 3.17% |
| Dec 2014 | 7 | 3.17% |
| Dec 2015 | 6 | 2.71% |
| Dec 2016 | 13 | 5.88% |
| Dec 2017 | 4 | 1.81% |
| Dec 2018 | 8 | 3.62% |
| Dec 2019 | 11 | 4.98% |
| Dec 2020 | 11 | 4.98% |
| Dec 2021 | 5 | 2.26% |
| Dec 2022 | 6 | 2.71% |
| Dec 2023 | 9 | 4.07% |
| Dec 2024 | 9 | 4.07% |
| Dec 2025 | 15 | 6.79% |
| Total | 221 | - |
Visualizing migrant deaths by December and year
Below, I produce a bar plot of migrant deaths by month for each year. Each bar corresponds to the number of migrant remains recovered in a given month for each year.
Remains recovered, sorted from lowest to highest, by month and year.
Making sense of the PMI
In the OpenGIS data, the Medical Examiner records the likely “post-mortem interval.” This is essentially the time frame in which a migrant likely died. Thus, the code “< 1 day” would imply the migrant died within the past 24 hours, while the code “> 6-8 months” means that the migrant likely died, potentially, much further back in time. Below is a table of the data. One way I like to think of the death data is in terms of what I call “contemporaneous deaths.” By this I mean deaths that likely occurred within a 12-month window. To explain, suppose a migrant remains are found in February? Suppose PCOME assesses the PMI as death occuring less than 6-8 months later. This would seem to imply that the migrant didn’t die in the year his/her death is recorded, but did die within a 12-month window. For those whose PMI is assessed as being greater than 6-8 months, it’s impossible to determine what the time window is in which they died. It could be one year or 15 years.
The following is a table of the PMI for a given month for all years.
December_data$Post Mortem Interval |
n | percent | valid_percent |
|---|---|---|---|
| < 1 day | 32 | 14% | 15% |
| < 1 week | 5 | 2% | 2% |
| < 3 months | 30 | 14% | 14% |
| < 3 weeks | 2 | 1% | 1% |
| < 5 weeks | 3 | 1% | 1% |
| < 6-8 months | 36 | 16% | 17% |
| > 6-8 months | 106 | 48% | 50% |
| NA | 7 | 3% | - |
Of interest might be to consider how the PMI codes vary with respect to time. Below I create a table to inspect this.
| year_month/pmi | < 1 day | < 1 week | < 3 months | < 3 weeks | < 5 weeks | < 6-8 months | > 6-8 months | NA_ | Total |
|---|---|---|---|---|---|---|---|---|---|
| Dec 2000 | 0% (0) | 100% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 100% (1) |
| Dec 2001 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 50% (1) | 50% (1) | 0% (0) | 100% (2) |
| Dec 2002 | 33% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 67% (2) | 0% (0) | 100% (3) |
| Dec 2003 | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 33% (2) | 67% (4) | 0% (0) | 100% (6) |
| Dec 2004 | 60% (3) | 0% (0) | 0% (0) | 0% (0) | 20% (1) | 20% (1) | 0% (0) | 0% (0) | 100% (5) |
| Dec 2005 | 27% (3) | 9% (1) | 0% (0) | 0% (0) | 0% (0) | 9% (1) | 45% (5) | 9% (1) | 100% (11) |
| Dec 2006 | 13% (1) | 0% (0) | 0% (0) | 0% (0) | 0% (0) | 38% (3) | 38% (3) | 13% (1) | 100% (8) |
| Dec 2007 | 13% (1) | 13% (1) | 13% (1) | 0% (0) | 0% (0) | 13% (1) | 38% (3) | 13% (1) | 100% (8) |
| Dec 2008 | 29% (5) | 0% (0) | 0% (0) | 6% (1) | 0% (0) | 18% (3) | 41% (7) | 6% (1) | 100% (17) |
| Dec 2009 | 31% (5) | 0% (0) | 6% (1) | 0% (0) | 0% (0) | 0% (0) | 63% (10) | 0% (0) | 100% (16) |
| Dec 2010 | 0% (0) | 0% (0) | 10% (1) | 0% (0) | 0% (0) | 40% (4) | 50% (5) | 0% (0) | 100% (10) |
| Dec 2011 | 8% (1) | 0% (0) | 25% (3) | 0% (0) | 0% (0) | 17% (2) | 42% (5) | 8% (1) | 100% (12) |
| Dec 2012 | 18% (2) | 9% (1) | 0% (0) | 0% (0) | 0% (0) | 27% (3) | 36% (4) | 9% (1) | 100% (11) |
| Dec 2013 | 0% (0) | 0% (0) | 14% (1) | 14% (1) | 0% (0) | 43% (3) | 29% (2) | 0% (0) | 100% (7) |
| Dec 2014 | 0% (0) | 0% (0) | 43% (3) | 0% (0) | 0% (0) | 14% (1) | 43% (3) | 0% (0) | 100% (7) |
| Dec 2015 | 17% (1) | 0% (0) | 17% (1) | 0% (0) | 0% (0) | 50% (3) | 17% (1) | 0% (0) | 100% (6) |
| Dec 2016 | 8% (1) | 8% (1) | 23% (3) | 0% (0) | 8% (1) | 8% (1) | 38% (5) | 8% (1) | 100% (13) |
| Dec 2017 | 0% (0) | 0% (0) | 25% (1) | 0% (0) | 0% (0) | 0% (0) | 75% (3) | 0% (0) | 100% (4) |
| Dec 2018 | 0% (0) | 0% (0) | 50% (4) | 0% (0) | 0% (0) | 13% (1) | 38% (3) | 0% (0) | 100% (8) |
| Dec 2019 | 9% (1) | 0% (0) | 18% (2) | 0% (0) | 0% (0) | 9% (1) | 64% (7) | 0% (0) | 100% (11) |
| Dec 2020 | 9% (1) | 0% (0) | 9% (1) | 0% (0) | 0% (0) | 9% (1) | 73% (8) | 0% (0) | 100% (11) |
| Dec 2021 | 0% (0) | 0% (0) | 40% (2) | 0% (0) | 0% (0) | 20% (1) | 40% (2) | 0% (0) | 100% (5) |
| Dec 2022 | 17% (1) | 0% (0) | 33% (2) | 0% (0) | 0% (0) | 0% (0) | 50% (3) | 0% (0) | 100% (6) |
| Dec 2023 | 22% (2) | 0% (0) | 22% (2) | 0% (0) | 11% (1) | 11% (1) | 33% (3) | 0% (0) | 100% (9) |
| Dec 2024 | 22% (2) | 0% (0) | 11% (1) | 0% (0) | 0% (0) | 11% (1) | 56% (5) | 0% (0) | 100% (9) |
| Dec 2025 | 7% (1) | 0% (0) | 7% (1) | 0% (0) | 0% (0) | 7% (1) | 80% (12) | 0% (0) | 100% (15) |
| Total | 13% (32) | 5% (5) | 14% (30) | 1% (2) | 1% (3) | 18% (36) | 45% (106) | 3% (7) | 100% (221) |
Below is a stacked barplot for each month by PMI and year.
To simplify our understanding of the data, consider this figure. Here I compare cases where the PMI is 5 weeks or less to cases that are: 3 months or less; 6-8 months or less; and cases greater than 6-8 months.
