This USA map shows wThis U.S. state map shows how actual 2020 deaths in each state compared with the expected number. It's s mostly pink. Arizona and New Jersey are red, meaning bad, and Mississiippi, Texas, CColorado, South Dakota, Illinois and some other states are dark pink. North Carolina and West Virginia look green, for some reason. (Credit: CDC)

Early government numbers show that at least 3.2 million people died in the United States in 2020, and that the number of deaths was at least 12% higher than expected.

The National Center for Health Centers — an arm of the U.S. Centers for Disease Control and Prevention (CDC) — has published those figures in a new batch of preliminary COVID-19 death count data. The batch covers death counts for the period extending from the week ending Jan. 4, 2020, to the week ending Jan. 2, 2021.

At least 315,936, or 9.8%, of all of the deaths in the data involved COVID-19.


The preliminary numbers also include about 28,000 excess deaths not classified as involving COVID-19.

Some of those deaths could have been the result of COVID-19 cases that were not diagnosed.

Other deaths could have been the result of the effects of the COVID-19 pandemic on patients’ access to high-quality health care. In the April, for example, some hospitals canceled or postponed elective procedures, in an effort to save capacity for patients with COVID-19 and to minimize the risk of patients infecting hospital workers, or of hospital workers infecting patients.

Some states take a month or more to send complete death count data to the CDC. The CDC appears to be missing about half of the death count data it will eventually receive for the period from Dec. 6 through Jan. 2. If the actual number deaths during that period was at least as a high as the number during the comparable period a year earlier, then 2020 death total could increase to about 3.3 million.

The Context

In January 2020, before public health officials knew what the effects of COVID-19 on the United States would be, executives at Reinsurance Group of America Inc. talked about the assumptions they use in models for “200-year event” pandemics.

RGA executives say they have been defining a 200-year pandemic as one that increases mortality by about 1.5 deaths per 1,000 lives.

In a normal year, the United States experiences about 8.7 deaths per 1,000 lives. That means RGA’s 200-year event pandemic would increase mortality about 17% over the expected level of mortality.

State Data

At the state level, in the preliminary data, ratios of actual deaths to expected deaths varied widely. In North Carolina, the actual number of deaths amounted to 80% of the expected number. In New Jersey, the state with the worst preliminary mortality figures, the actual number of deaths amounted to 125% of the expected number.

The National Center for Health Statistics breaks out figures for New York City separately from the figures for New York state. In New York.

Officials have noted that the numbers for North Carolina and West Virginia appear to be below average mainly because those states have been very slow to report data. North Carolina has not sent in complete death count data since August 2020. The latest numbers for North Carolina, for example are for the week ending Oct. 1.

From January through March, the actual number of deaths in North Carolina was only about 1% over the expected number.

From April through August, the actual number of deaths was about 10% higher than expected.

For all months included in the North Carolina data, the actual number of deaths was 7% higher than expected.

National Center for Health Statistics reports are in the public domain. That means insurers, actuaries, insurance agents and other financial professionals can use the data in their own reports and publications without worrying about copyright restrictions.

Government analysts calculate the “expected number of deaths” by taking the average of the number of deaths that occurred in a given week or month during the previous three years. For January 2020, for example, the expected number of deaths was equal to the average of the death counts for January 2017, January 2018 and January 2019.

For preliminary state-level actual-to-expected death count ratios, see the table below.


Preliminary 2020 U.S. Deaths

Actual Deaths as a Percent of Expected Deaths
Alabama 112
Alaska 103
Arizona 122
Arkansas 113
California 111
Colorado 116
Connecticut 106
Delaware 108
District of Columbia 114
Florida 113
Georgia 113
Hawaii 102
Idaho 111
Illinois 116
Indiana 111
Iowa 112
Kansas 112
Kentucky 105
Louisiana 114
Maine 105
Maryland 115
Massachusetts 112
Michigan 114
Minnesota 112
Mississippi 118
Missouri 111
Montana 111
Nebraska 110
Nevada 114
New Hampshire 107
New Jersey 125
New Mexico 112
New York 7 116
New York City 149
North Carolina 80
North Dakota 113
Ohio 107
Oklahoma 107
Oregon 103
Pennsylvania 108
Rhode Island 105
South Carolina 116
South Dakota 116
Tennessee 113
Texas 117
Utah 112
Vermont 101
Virginia 112
Washington 106
West Virginia 91
Wisconsin 113
Wyoming 114
Puerto Rico 91
United States  112

Clarification: An earlier version of this article gave no explanation of why the death numbers in North Carolina and West Virginia are so low. The death numbers for those states are low partly because of data submission timing.

— Read How Likely the Government Thinks You Are to Die This Year, on ThinkAdvisor.

— Connect with ThinkAdvisor Life/Health on FacebookLinkedIn and Twitter.