COVID-19 short-term forecasts Deaths 2021-08-04


General information

  • Forecasts produced by Jennie Castle, Jurgen Doornik, and David Hendry, researchers at the University of Oxford. These are our forecasts, and should not be considered official forecasts from, or endorsed by, any of: University of Oxford, Oxford Martin School, Nuffield College, or Magdalen College.
  • These forecasts are short term time-series extrapolations of the data. They are not based on epidemiological modelling or simulations. All forecasts are uncertain: their success can only be determined afterwards. Many mitigation strategies are in place, which, if successful, invalidate our forecasts. An explanation of our methods is provided below.
  • A list of notes is below. The most recent note:
    [2021-04-29]The `legacy' download for areas of England is stuck at April 26, so we switched to the newer downloads. The results now include Scotland, Wales, and Northern Ireland. The map, however, only shows England.

Peak increase in estimated trend of Deaths in Europe 2021-08-04

UKEUATBEBGBSCZDEDKESFIFRGRHRHUIEITNLPLPTROSESISKNOCH
Peak date (mm-dd)2021-01-1911-2612-1711-032021-04-0712-282021-03-092021-01-132021-01-1511-0412-2111-162021-05-032021-04-282021-03-312021-01-1912-012021-01-082021-04-132021-01-282021-06-082021-01-2012-042021-04-072021-04-2111-16
Peak daily increment 1259 3528 112 202 124 55 215 857 38 927 6 601 82 44 253 56 724 99 540 279 173 139 52 105 11 96
Days since peak 197 251 230 274 119 219 148 203 201 273 226 261 93 98 126 197 246 208 113 188 57 196 243 119 105 261
Last total 130000 743511 10744 25258 18229 6519 30372 91736 2550 81844 984 110952 12988 8267 30032 5044 128136 17839 75269 17412 34298 14657 4429 12541 804 10908
Last daily increment 119 217 2 7 4 4 -6 26 0 71 0 30 16 0 3 9 21 7 4 15 1 0 0 0 5 2
Last week 485 1187 7 23 21 12 9 114 2 402 2 212 77 17 7 9 107 24 12 82 23 1 1 5 5 3
Previous peak date04-1004-0404-0604-1512-02 --11-0404-1504-0403-3104-2104-0511-2912-1912-1104-2403-2904-0411-2704-1511-2704-21 -- -- --04-06
Previous peak daily increment 964 3092 20 328 142 207 226 15 946 30 933 101 76 164 167 781 153 492 31 164 101 60
Low between peaks 8 -751 0 -106 35 97 3 0 -1351 0 7 23 10 80 0 6 -6 204 3 50 1 -54

Peak increase in estimated trend of Deaths 2021-08-04

AustraliaBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-AKUS-ALUS-ARUS-AZUS-CAUS-COUS-CTUS-DCUS-DEUS-FLUS-GAUS-HIUS-IAUS-IDUS-ILUS-INUS-KSUS-KYUS-LAUS-MAUS-MDUS-MEUS-MIUS-MNUS-MOUS-MSUS-MTUS-NCUS-NDUS-NEUS-NHUS-NJUS-NMUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-SDUS-TNUS-TXUS-UTUS-VAUS-VTUS-WAUS-WIUS-WVUS-WY
Peak date (mm-dd)09-032021-04-062021-01-162021-07-042021-06-252021-06-10 --2021-04-252021-06-012021-04-112021-04-092021-07-292021-07-292021-04-272021-01-152021-04-152021-02-0212-142021-01-152021-02-2412-092021-01-112021-02-032021-07-302021-01-212021-01-252021-01-262021-04-0312-1012-1012-162021-07-222021-06-0111-1608-202021-05-272021-03-0912-072021-03-092021-05-182021-01-142021-04-092021-02-0212-0912-0305-2906-252021-05-262021-01-132021-01-1412-162021-04-072021-04-062021-01-132021-02-0112-032021-06-112021-01-192021-06-252021-03-032021-01-0212-292021-07-282021-03-122021-01-07
Peak daily increment 35 2996 147 119 654 6246 420 3098 813 223 778 675 354 3330 14 134 43 154 612 151 38 5 130 178 138 48 54 18 148 100 13 230 155 106 480 8 126 114 71 45 14 96 27 25 25 1566 94 41 191 194 1325 20 211 94 23 18 336 9 326 2 80 72 128 10
Days since peak 335 120 200 31 40 55 101 64 115 117 6 6 99 201 111 183 233 201 161 238 205 182 5 195 191 190 123 237 237 231 13 64 261 349 69 148 240 148 78 202 117 183 238 244 432 405 70 203 202 231 119 120 203 184 244 54 197 40 154 214 218 7 145 209
Last total 932 559607 26582 35671 121695 426290 100636 92194 242547 196673 28231 159032 73415 51767 614785 393 11561 6230 18289 64620 6963 8291 1149 1833 36973 21028 535 6193 2210 23477 13610 5264 7362 11129 18074 9814 900 21042 7683 9593 7597 1718 13700 1541 2230 1388 26625 4416 5953 53405 20526 7531 2877 27883 9930 2052 12588 53457 1683 11541 260 6142 8302 2956 786
Last daily increment 5 1175 13 31 479 533 1747 409 611 75 90 769 423 122 490 2 19 15 7 22 6 0 0 0 0 11 0 10 1 18 14 3 10 44 0 4 0 0 5 26 7 2 21 1 0 1 11 1 24 3 0 46 5 15 13 7 10 62 3 5 0 9 3 2 0
Last week 9 5110 38 376 1569 3073 10084 2120 2550 459 654 4628 1984 583 2661 7 51 120 89 234 28 -1 0 133 0 66 4 10 15 44 40 11 35 161 7 19 1 52 20 138 64 14 82 3 0 2 35 8 53 44 38 46 22 41 26 7 53 271 14 16 1 26 15 12 10
Previous peak date --07-2104-2907-172021-01-2206-162021-01-2711-1810-0507-1809-14 --2021-01-1412-2604-132021-01-2009-1609-1507-2108-0604-2404-2004-2904-2708-0711-0310-2112-1109-0905-1205-022021-01-132021-03-1804-1404-2504-292021-01-0804-1412-1112-1508-262021-02-03 -- --05-20 --04-2212-1608-1305-1804-232021-01-2912-1505-0507-2807-292021-02-0507-2712-1004-2204-0904-0112-082021-01-1309-02
Previous peak daily increment 1065 186 785 389 1661 322 473 1724 918 159 566 253 2233 13 99 124 78 139 72 105 11 13 175 354 6 313 19 112 99 209 353 64 459 56 21 137 63 61 31 27 40 317 31 18 3552 43 40 23 135 45 2 113 461 9 104 11 60 60 26 27
Low between peaks 371 3 37 96 92 71 175 95 16 43 63 528 0 9 10 6 41 2 1 0 0 45 32 0 6 3 15 8 1 8 10 -1 5 1 6 8 -2 11 1 1 52 -85 5 7 12 1 2 12 14 1 5 70 1 9 0 6 -50 4 0

Deaths count forecast Europe (bold red line in graphs) 2021-08-05 to 2021-08-11

DateUKEUBEBGDEESFRGRITNLPTRO
2021-08-04 130000 743511 25258 18229 91736 81844 110952 12988 128136 17839 17412 34298
2021-08-05 130000 743500 25260 18230 91770 81960 111000 13000 128100 17840 17420 34300
2021-08-06 130000 743600 25260 18240 91800 82040 111000 13000 128200 17840 17420 34300
2021-08-07 130100 743700 25260 18240 91810 82090 111000 13000 128200 17840 17430 34310
2021-08-08 130100 743700 25260 18240 91810 82140 111000 13010 128200 17850 17440 34310
2021-08-09 130100 743900 25260 18250 91820 82230 111000 13010 128200 17850 17450 34310
2021-08-10 130300 744100 25260 18250 91840 82290 111100 13020 128200 17850 17460 34320
2021-08-11 130300 744300 25260 18250 91860 82330 111100 13030 128200 17860 17470 34320

Deaths count forecast (bold red line in graphs) 2021-08-05 to 2021-08-11

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-ARUS-AZUS-CAUS-COUS-DEUS-GAUS-ILUS-INUS-KYUS-LAUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-TNUS-TXUS-WA
2021-08-04 559607 26582 35671 121695 426290 100636 92194 242547 196673 28231 159032 73415 51767 614785 11561 6230 18289 64620 6963 1833 21028 23477 13610 7362 11129 21042 7683 9593 7597 13700 26625 5953 53405 20526 7531 2877 27883 9930 12588 53457 6142
2021-08-05 561300 26590 35820 122000 426800 103800 92510 242500 196800 28320 159800 73750 51840 614800 11560 6230 18290 64620 6968 1838 21030 23480 13610 7362 11170 21070 7683 9593 7597 13700 26630 5964 53410 20530 7531 2879 27880 9934 12600 53460 6153
2021-08-06 562600 26600 35900 122400 427400 106500 92800 242600 196900 28390 160500 74030 51840 615300 11560 6230 18300 64630 6976 1842 21040 23490 13610 7368 11210 21110 7686 9593 7597 13700 26630 5971 53410 20530 7536 2882 27890 9934 12600 53490 6155
2021-08-07 563600 26610 35980 122700 427900 109100 93110 243000 197000 28470 161200 74340 51840 615300 11580 6233 18320 64630 6978 1846 21040 23490 13610 7369 11230 21130 7686 9600 7597 13700 26640 5971 53420 20530 7537 2882 27890 9934 12600 53510 6157
2021-08-08 564100 26610 36060 123000 428400 111300 93460 243100 197100 28550 161900 74540 51850 615300 11580 6240 18320 64630 6978 1849 21040 23490 13610 7370 11250 21150 7686 9600 7597 13700 26640 5971 53430 20530 7541 2882 27890 9934 12600 53520 6158
2021-08-09 564600 26610 36140 123300 428900 113600 93840 243200 197200 28630 162600 74730 51890 615600 11580 6266 18320 64700 6987 1853 21070 23500 13610 7385 11270 21180 7689 9600 7597 13720 26650 5974 53430 20530 7549 2886 27890 9934 12600 53540 6165
2021-08-10 565700 26620 36150 123600 429400 115900 94220 243800 197300 28720 163300 75170 51950 616200 11590 6280 18340 64730 6988 1856 21090 23510 13630 7391 11310 21220 7692 9671 7605 13730 26650 5984 53440 20530 7551 2895 27900 9934 12610 53590 6173
2021-08-11 566800 26630 36180 123900 429800 117600 94550 244300 197300 28800 164000 75460 52010 616600 11600 6291 18350 64760 6994 1860 21100 23520 13640 7400 11320 21220 7697 9693 7612 13750 26660 6007 53440 20540 7615 2899 27920 9937 12610 53650 6180

Deaths count average forecast Europe (bold black line in graphs) 2021-08-05 to 2021-08-11

DateUKEUBEBGDEESFRGRITNLPTRO
2021-08-04 130000 743511 25258 18229 91736 81844 110952 12988 128136 17839 17412 34298
2021-08-05 130100 743700 25260 18230 91750 81930 111000 13000 128200 17840 17420 34300
2021-08-06 130200 743900 25260 18230 91760 81990 111000 13010 128200 17850 17430 34300
2021-08-07 130200 744000 25260 18240 91760 82030 111100 13010 128200 17850 17440 34300
2021-08-08 130300 744100 25260 18240 91770 82080 111100 13020 128200 17850 17450 34300
2021-08-09 130300 744300 25270 18240 91780 82140 111100 13030 128200 17850 17460 34310
2021-08-10 130400 744500 25270 18240 91790 82200 111100 13040 128200 17860 17470 34310
2021-08-11 130500 744600 25270 18240 91820 82250 111100 13050 128200 17860 17480 34310

Deaths count average forecast (bold black line in graphs) 2021-08-05 to 2021-08-11

DateBrazilCanadaChileColombiaIndiaIndonesiaIranMexicoPeruPhilippinesRussiaSouth AfricaTurkeyUSUS-ALUS-ARUS-AZUS-CAUS-COUS-DEUS-GAUS-ILUS-INUS-KYUS-LAUS-MIUS-MNUS-MOUS-MSUS-NCUS-NJUS-NVUS-NYUS-OHUS-OKUS-ORUS-PAUS-SCUS-TNUS-TXUS-WA
2021-08-04 559607 26582 35671 121695 426290 100636 92194 242547 196673 28231 159032 73415 51767 614785 11561 6230 18289 64620 6963 1833 21028 23477 13610 7362 11129 21042 7683 9593 7597 13700 26625 5953 53405 20526 7531 2877 27883 9930 12588 53457 6142
2021-08-05 560700 26590 35770 122000 427000 102500 92490 243000 196800 28310 159800 73830 51850 615100 11570 6243 18300 64670 6970 1835 21040 23490 13620 7367 11170 21050 7687 9613 7608 13710 26630 5970 53410 20540 7536 2881 27890 9936 12600 53490 6153
2021-08-06 561700 26600 35830 122300 427800 104200 92660 243300 196800 28390 160500 74160 51890 615500 11570 6250 18310 64690 6977 1839 21050 23490 13620 7372 11190 21070 7690 9621 7616 13720 26630 5983 53420 20540 7538 2884 27890 9938 12600 53530 6155
2021-08-07 562600 26610 35900 122500 428500 105800 92830 243700 196900 28480 161200 74480 51940 615500 11580 6258 18320 64700 6980 1843 21060 23490 13620 7374 11190 21070 7691 9631 7618 13730 26640 5987 53420 20540 7539 2885 27890 9940 12600 53550 6156
2021-08-08 563200 26610 35950 122800 429200 107300 93010 243800 197000 28560 161900 74760 51990 615500 11580 6266 18330 64700 6982 1846 21060 23490 13620 7376 11200 21080 7691 9637 7621 13730 26640 5990 53430 20550 7540 2886 27900 9942 12610 53550 6156
2021-08-09 563900 26610 36010 123000 429800 108700 93210 244000 197100 28650 162600 75050 52050 615800 11580 6276 18330 64760 6989 1849 21070 23500 13620 7385 11220 21090 7694 9644 7628 13740 26640 5995 53430 20550 7542 2890 27900 9945 12620 53570 6159
2021-08-10 565000 26620 36040 123200 430300 110300 93400 244400 197200 28740 163400 75440 52110 616000 11580 6284 18340 64780 6991 1852 21090 23510 13630 7389 11240 21100 7696 9677 7640 13750 26640 6006 53440 20560 7543 2895 27910 9947 12620 53600 6164
2021-08-11 566100 26630 36060 123400 430900 111800 93560 244900 197300 28830 164100 75700 52170 616200 11590 6294 18350 64800 6995 1855 21100 23510 13640 7393 11250 21110 7699 9690 7647 13760 26650 6024 53440 20560 7566 2898 27910 9951 12630 53630 6167

Further information

  • We believe these forecasts fill a useful gap in the short run. They give an indication of what is likely to happen in the next few days, removing some aspect of surprise. Moreover, a noticeable drop in comparison to the extrapolations could be an indication that the implemented policies are having some impact. It is difficult to understand exponential growth. We hope that these forecasts may help to convince viewers to adhere to the policies implemented by their respective governments, and keep all arguments factual and measured.
  • We use the data repository for the 2019 Novel Coronavirus Visual Dashboard operated by the Johns Hopkins University Center for Systems Science and Engineering. This is updated daily, but we tend to update our forecasts only every other day.
    US state data as of 2020-03-28 is courtesy of the New York Times.
  • We can only provide forecasts of what is measured. If confirmed cases are an underestimate of actual cases, then our forecasts will also be underestimates. No other epidemiological data is used. Data definition and collection differs between countries and may change over time.
  • We will update the methodology as we learn what is happening in the next few days or weeks. Once the number of cases levels off, there is no need to provide these forecasts anymore.
  • Countries where the counts are very low or stable have been omitted.
  • The graphs have dates on the horizontal axis (yyyy-mm-dd) and cumulative counts on the vertical axis. They show
    1. bold dark grey line (with circles): observed counts (Johns Hopkins CSSE);
    2. many light grey lines (with open circles): forecasts using different model settings and starting up to four periods back;
    3. red line (with open circles): single forecasts path using default model settings;
    4. black line (with crosses): average of all forecasts, recentered on the last observation;
    5. thin green lines: some indication of uncertainty around the red forecasts, but we do not know how reliable that is.
    Both the red line forecasts and the black lines are also given in the tables above. These forecasts differ, we are currently inclined to use the average forecasts.
  • The forecasts are constructed as follows:
    1. An overall `trend' is extracted by taking a window of the data at a time. In each window we draw `straight lines' which are selected using an automatic econometric procedure (`machine learning'). All straight lines are collected and averaged, giving the trend.
    2. Forecasts are made using the estimated trend, but we note that this must be done carefully, because simply extrapolating the flexible insample trend would lead to wildly fluctuating forecast. We use the `Cardt' method, which has been found to work well in other settings.
    3. Residuals from the trend are also forecast, and combined with trend forecasts into an overall forecast.
  • Scenario forecasts are constructed very differently: smooth versions of the Chinese experience are matched at different lag lengths with the path of each country. This probably works best from the peak, or the slowdown just before (but we include it for the UK nonetheless).
  • The forecast evaluation shows past forecasts, together with the outcomes (in the grey line with circles).
  • EU-BS is Estonia, Latvia, and Lithuania together.
  • This paper describes the methodology and gives further references. Also available as Nuffield Economics Discussion Paper 2020-W06. Still preliminary is the documentation of the medium term forecasts.

Recent changes and notes

[2021-04-29]The `legacy' download for areas of England is stuck at April 26, so we switched to the newer downloads. The results now include Scotland, Wales, and Northern Ireland. The map, however, only shows England.
[2021-01-07]Slideshow of forecasts, errors, and actuals 2020-06-30 to 2021-01-02: how England lost the battle.
[2020-10-27]Statistical short-term forecasting of the COVID-19 Pandemic (Jurgen Doornik, Jennie Castle, and David Hendry) is now published at the Journal of Clinical Immunology and Immunotherapy. open access
[2020-10-11]Short-term forecasting of the coronavirus pandemic (Jurgen Doornik, Jennie Castle, and David Hendry) is now in press at the International Journal of Forecasting. open access
[2020-10-10]Removed forecasts from the Chinese scenarios, while investigating possibility to use own history from the first wave.
Added information on the previous peak (if present) to the peak tables.
Local forecasts for England: now dropping last four observations.
[2020-07-01] Modified the short-term model to allow for (slowly changing) seasonality. Many countries show clear seasonality after the initial period, likely caused by institutional factors regarding data collection. This seasonality was also getting in the way of peak detection. As a consequence estimates of the peak date may have changed for countries with strong seasonality.
Added forecasts of cumulative confirmed cases for lower tier local authorities of England. The data is available from 2020-07-02 including all tests (pillar one and two). Only authorities with more than 5 cases in the previous week are included.
[2020-06-29] Tables in April included the world, but not the world as we know it (double counting China and the US). So removed the world from those old tables.
Why short-term forecasts can be better than models for predicting how pandemics evolve just appeared at The Conversation.
Thursday 2 July webinar at the FGV EESP - São Paolo School of Economics. This starts at 16:00 UK time (UTC+01:00) and streamed here.
[2020-06-24] Research presentation on short-term COVID-19 forecasting on 26 June (14:00 UK time) at the Quarterly Forecasting Forum of the IIF UK Chapter.
[2020-06-06] Removed Brazil from yesterday's forecasts (only; last observation 2020-06-05).
[2020-06-04] Data issues with confirmed cases for France.
Added an appendix to the short term paper with further forecast comparisons for European and Latin American countries.
Both Sweden and Iran have lost their peak in confirmed cases. For Sweden the previous peak was on 24 April (daily peak of 656 cases), for Iran it was on 31 March (peak of 3116). For Iran this looks like a second wave, with increasing daily counts for the last four weeks. For Sweden this is a sudden jump in confirmed cases in the last two days, compared to a fairly steady weekly pattern over the previous six weeks.
[2020-05-20] Problem with UK confirmed cases: negative daily count. This makes the forecasts temporarily unreliable.
Updated the second paper.
[2020-05-18] Minor fixes to the improved version of scenario forecasting, backported to 2020-05-13.
[2020-05-13] We now omit countries with fewer than 200 confirmed cases in the last week (25 for deaths).
The short-term paper has some small updates, including further comparisons with other models.
Data for Ecuador are not reliable enough for forecasting.
Switched to an improved version of scenario forecasting.
[2020-05-06] The New York Times is in the process of redefining its US state data. Unfortunately, at the moment only the last observation has changed (e.g New York deaths jumped from 19645 on 2020-05-05 to 25956 a day later). This means the data is currently useless; however it does bring it close to the Johns Hopkins/CSSE count (25626 on 2020-05-06). The aggregate US count is based on JH/CSSE so unaffected. We now use Johns Hopkins/CSSE US state data, including all states with sufficient counts. So the new forecasts cannot be compared to those previously.
A minor change is that we show the graph without scenario forecast if no peak has been detected yet.
[2020-04-29] See our blog entry at the International Institute of Forecasters.
US history of death counts revised in Johns Hopkins/CSSE data.
UK death counts have been revised to include the deaths in care homes. In the Johns Hopkins/CSSE data set, which we use, the entire history has been revised. So forecasts made up to 2020-04-29 cannot be compared to later outcomes. In the ECDC data set only the last observation has changed, causing a jump in the series.
[2020-04-27] Our short-term COVID-19 forecasting paper is now available as Nuffield Economics Discussion Paper 2020-W06.
A small adjustment has been made to the scenario forecast methodology, and will be documented shortly.
[2020-04-24] A summary of our work on short-term COVID-19 forecasting appeared as a voxeu.
[2020-04-17] Bird and Nielsen look into nowcasting death counts in England.
[2020-04-16] Added scenario forecasts to all graphs now. This would now be the preferred forecast for most.
This is the first time with a peak in confirmed UK cases (also for deaths, but this is uncertain because it is at the same date).
[2020-04-10] Updated documentation with better description of short-term estimates and peak determination.
[2020-04-09] Added table with estimated peak dates (if happened) and dates to and since the peak. Note that this can be a local peak, and subsequent re-acceleration (or data revisions) can result in a new peak later.
[2020-04-08] Minor correction to peak estimates. Added table with scenario forecasts.
[2020-04-06] Added a post hoc estimate of the peak number of cases. This needs at least three confirmed observations (four for deaths) after the event. It is based on the averaged smooth trend, and can change later or be a local peak. It is marked with a vertical line with the date label, or a date with left arrow in the bottom left corner of the graph. This is backported to 2020-04-04.
[2020-04-02] Now including more US States, based on New York Times data.
[2020-03-31] Scenario forecasts, based on what happened in China earlier this year, are presented for several countries (line marked with x). Created more plausible 90% confidence bands (dotted line in same colour).
[2020-03-26] Scenario forecasts that are based on what happened in China earlier this year, only for Italy.
[2020-03-24] Our forecasts are starting to overestimate in some cases. This was always expected to happen when the increase starts to slow down. Scenario forecasts that are based on what happened in China earlier this year, but only for Italy and Spain sofar.

Initial visual evaluation of forecasts of Deaths