You are on page 1of 16

Declaration of Raymond M.

Pursuant to 28 U.S.C Section 1746, I, Raymond M. Blehar, make the following

1. I am over the age of 21 years and I am under no legal disability, which would
prevent me from giving this declaration.

2. I am a resident of Anne Arundel County, Maryland.

3. I earned a Masters of Business Administration degree (with honors) from Penn

State University in 2008 and a Bachelor of Arts degree in Geography from the
Indiana University of Pennsylvania in 1983. I retired from the federal
government in 2017 after serving 32 years in the Defense/Intelligence
Community. While employed by the National Geospatial Intelligence Agency
(NGA) and predecessor agencies (1985 to 2017), I served in a variety of
management and staff positions including (but not limited to) assistant inspector
general, performance analyst, primary staff officer for intelligence program
reviews, head of financial management control, financial audit liaison to the
Department of Defense (DOD), and assistant deputy director for quality. My
experiences at NGA included various statistical analysis and modeling activities,
administration and analysis of organizational surveys, and monitoring and
analyzing agency performance to include financial analysis, audit, and oversight.
From 2010 to 2013, I served a rotational assignment as a senior analyst to the
Office of the Director of National Intelligence (ODNI), Systems, Resources, and
Analysis Directorate. While at ODNI, I conducted various analyses regarding
multiple systems software and hardware acquisitions in excess of $10 billion
(aggregate). In addition, I provided financial analyses and recommendations
regarding investments in commercial partnerships. During my career I
received commendations from NGA (and predecessors), ODNI, CIA, and the
Department of Defense for my analytic contributions.

4. In addition to my work activities, I volunteered as an Awards Examiner for the
U.S. Senate Productivity Awards and Maryland Quality Awards program (1997-
2001). In that role, I evaluated the management practices of public agencies,
private sector companies, and educational systems and provided
recommendation s for improvement. More recently (2012-2018), I provided pro
bono investigative/analytic support to defense teams in the criminal cases of
former Penn State President Graham Spanier (Commonwealth v. Spanier) and
former Pennsylvania Attorney General Kathleen Kane (Commonwealth v. Kane)
as well as to Spanier as the plaintiff in his civil case against former Federal
Bureau of Investigation (FBI) Director, Louis Freeh (Spanier v. Freeh). Notably,
my investigations/analyses revealed that the key evidence used in the
prosecutions/convictions was of dubious provenance and was subjected to

5. My analysis of the data relating to Pennsylvania’s 2020 general election
concludes that there were numerous anomalies and/or inconsistencies in the
results and that an estimated 280,000 – 300,000 Democrat, Republican,
Libertarian, and Green Party electors (voters) were disenfranchised because
their votes were manipulated (i.e., changed) by the election system software,
election administrators, and/or by introduction of other disruptive software (i.e.,
malware) or a combination of the three.

6. Like many people in the United States of America (U.S.) and across the world, I
watched the lead up to the 2020 U.S. election with great interest. During that
time, I learned that many U.S. states extended the deadlines for returning
ballots. As such, on November 3, 2020, I did not expect the election would be
finalized and believed that it could take into the following week to decide the

winner. On November 4, 2020, I checked the results for the key battleground
state of Pennsylvania at approximately 11:00 AM EST and learned that
President Trump held an approximate lead of 522,000 votes1 with less than 1.4
million votes left to count. I quickly did the back of the envelope math and
determined that Biden would need to win over 70% of the remaining vote to
overtake Trump.

Figure 1: Trump winning by 10% at 10:35 AM with over 1 million votes to count

7. Given that the vote count on Wednesday until the end of the count would consist
of mail-in votes, I checked the Pennsylvania Secretary of State web-site for the
statistics on mail-in ballots returned by political party. The figures showed the
proportions of returned ballots were approximately 64.8% Democrat, 23.7%
Republican, 10.7% No Party Affiliation, and .08% Minor Party. As such, it was
highly unlikely that Biden could win even if he got 90% of the remaining
Democrat votes, a 10% cut of GOP votes, and a 60% share of No Party Affiliation
(i.e., independent) voter. Under that very generous scenario,2 Trump would
have won Pennsylvania by nearly 43K votes. My projection follows at Figure 2.

I made a cryptic note of the vote count at 11:00AM and it stood Trump 3027K to Biden 2505K.
Exit polling had Biden at 52% of Independents I increased the percentage to 60% for the analysis.

11:00 AM EST Projected Total Margin of Victory
Trump 3,027,000 460,460 3,487,460 42,920
Biden 2,505,000 939,540 3,444,540

Figure 2: Projection of PA Election based on votes remaining and polling data

8. Biden’s eventual win by over 80,000 votes with 6.9M votes cast did not appear to
be possible based on the votes outstanding by party affiliation and voter

9. In the days following the announcement that Biden won Pennsylvania by nearly
80,000 votes, a blogger3 alleged 220,000 votes were shifted from Trump to Biden
in Pennsylvania based on an analysis of Edison Research’s election results and
posted the underlying source data on the blog. I accessed the Edison Data
scraped from the New York Time’s (NYT) web-site4 to perform an analysis and
determine if the claim was legitimate.

Analysis of Alleged Switching of 220,000 Votes
10. I confirmed the blogger’s outcome of 220,000 switched Trump votes after
downloading the Edison/NYT data, importing it to Excel, and performing
analyses. However, because I used a later date file with more election returns,
my analysis revealed that approximately a 280,000 votes were switched
including Biden and third party votes. A sample of the Edison/NYT batch files



11. Note that the sample files include only percentages to two or three decimal
places for Trump and Biden and the total number of votes cast. Determining the
votes for the “Third Party”5 was done through computation. Therefore, the
results of the analysis should be considered as preliminary.6 That said, there
were many instances where the “Third Party” votes were of no consequence, but
showed high numbers of vote switches directly between the two candidates.
Examples follow:

12. In the following example (Figure 3.) only 54 New Votes were recorded at that
instant, however Biden gained an impossible 17,930 votes. Trump’s 17,877 were
moved to Biden to make up the difference. Other vote swaps found in the
analysis ranged from approximately 1,000 to 7,000 votes, proportional to the
total number of votes cast. As such, this large swap of 17,877 is likely from
human intervention.

Figure 3: Vote switch from Trump to Biden

13. Two instances of vote switching from Trump to Biden are demonstrated in the
on the next page (Figure 4).

Third Party includes ballots/votes from Libertarian and Other voters (Green Party is included in other).
Examination of electronic voting data from voting systems is required to verify the analysis

Figure 4: Vote Switching Example: Trump to Biden

14. In the top row, New Votes are 6,669 of which the “Third Party” received 87. In
order to reach 8,803 votes, Biden received 2,220 votes from Trump. Math: 6,669
(N) + 2,220 (T) – 87 (G) = 8,802 (B) In the bottom example, New Votes are 5,392
of which the “Third Party” received 70. In order to reach 8,417 votes, Biden
received 3,095 votes from Trump. Math: 5,392 (N) + 3095 (T) – 70 (G) = 8417

Figure 5: Vote Swap: Biden to Trump

15. A vote swap from Biden to Trump that occurred early in the vote counting is
depicted in Figure 4. In this transaction, Trump received 4,813 votes – 356

more votes than the 4,420 New Votes that came in. Biden votes made up the
difference. Math: 4,520 (N) – 63 (G) + 356 (B) = 4,813 Trump.

16. Preliminary analysis found many instances of “Third Party” vote laundering.
In other words, votes were moved from Trump or Biden to the “Third Party” then
allocated to either Trump or Biden in a later transaction. Again, “Third Party”
does not mean that votes were moved to Jorgenson. Rather votes were likely
moved to an adjudication folder for later reassignment. An example of “Third
Party” vote laundering is shown at Figure 6.

Figure 6: Third Party vote laundering example

17. Starting at the top, 6,364 votes are moved from “Third Party” to Biden. The
math is: 9,526 (N) – 4,773 (Trump) + 6,364 (Third) = 11,1177 (Biden.) The next
highlighted line shows Trump losing 4,322 votes to the “Third Party.” The math
is 4,416 (N) – 2,155 (B) = 2,261 (Third) + 4,322 (Trump) = 6,583 (Third). The next
transaction moves 6,385 votes from the Third Party to Biden. The math is
12,480 (N) – 6,228 (Tr) = 6,252 (B) + 6,385 (Third) = 12,637 (Biden). The next
transaction shows Trump losing 6,335 votes to the Third Party. The math goes
426 (N) – 208 (B) = 218 (Third) + 6,335 (Trump) = 6, 553 (Third). I refer to this

Off by 1 due to rounding of decimal places.

technique as “vote laundering” because it’s not a direct transfer. I refer to the
series of transactions as the “pump and dump” because it pumps votes to the
“Third Party” then dumps the votes in a later transaction.

18. Figure 7 provides a set of captioned graphics that demonstrate the “pump and
dump” after 76 percent of the Pennsylvania votes were counted; the critical time
that Biden overtook Trump.

Figure 7: The "pump and dump" after 76 percent

19. Quantitative results related to vote switching follow:
a. Biden received votes directly from Trump on 9 occasions for 31,697 votes.
b. Trump received 2,481 votes directly from Biden over 4 transactions.
c. Biden received votes from the “Third Party” 45 times for a grand total of
205,237 added votes. Another 19,732 votes that were allocated from his
total to the “Third Party” were recycled back to him.
d. Trump received 30,155 votes via transfer from the “Third Party.”

20. It bears repeating, this analysis is preliminary based on the Edison Research
data scraped8 from the NYT. Examining the voting system is the only way to
determine if these transactions occurred.

Scenario Analysis Supports Switch of 220,000 votes

21. I also performed scenario analysis of the Pennsylvania presidential results
using mail-in vote return statistics and exit polling data.9 The scenario that
most closely resembled exit polling data produced a result of Trump receiving
215,311 more mail-in votes. The full analysis is at Appendix A.
Scenario 3: TOTAL MAIL IN VOTES and TOTAL VOTES have freedom of movement
Polling factors applied to Base,Other, and Disloyals
Outside REVISED Disloyal Outside Outside
REVISED Mail In Ballots Polling Same Party Party MAIL IN Disloyal Party Vote obtained
TOTAL VOTES Cast By Party Loyalty Support Support VOTES PCT Votes Available PCT
BIDEN 3,202,811 (D) 1,711,192 92% 1,574,297 204,548 1,778,845 8% 136,895 336,951 61%
TRUMP 3,553,332 ( R) 627,302 91% 570,845 234,039 804,884 9% 56,457 417,389 56%
JORGENSON 78,238 (L) 8,136 100% 8,136 16,360 24,496 0% 0 490,207 3%
Other O 296,854 0 0
Missing Mail In [32,259]
TOTAL 6,834,381 2,643,484 2,608,225 0 193,353 Deltas
Trump 44% of Other 87.50% of Disloyal Democrats Biden -189,547
Biden 52% of Other 88.89% of Disloyal Republicans Trump 215,311

Figure 8: Trump receives 215K more mail-in votes than official results under exit polling scenario

Data scraping refers to a technique in which a computer program extracts data from output generated from
another program. Data scraping is commonly manifest in web scraping, the process of using an application to
extract valuable information from a website.

Other Pennsylvania Voting Information that Support Vote Switching
22. There are visible inconsistencies in the comparison of party voting patterns for
the four state-wide candidates that also suggest vote switching occurred on the
order of 20,000 – 30,000 votes being reassigned from the Green Party to Biden.

Inconsistencies: Third Party Voting Patterns

23. Green Party voters had to write-in their Presidential candidate. As such, their
votes for President were recorded as “Other” in the voting system. Write-in
votes required adjudication where an election worker, under supervision,
reassigned the vote. Given that a large number of write-in votes were expected
in advance, software could be loaded to the system to reassign a portion of those
votes without going through adjudication.

24. Between 71K and 80K people voted for the Green Party candidates for Attorney
General, Auditor General, and Treasurer. Those votes seemingly disappeared
in the Presidential race as only 8,000 write in votes were recorded10 in that
contest. The uniformity of the Green Party votes across the board for AG, Auditor
General, and Treasurer is unusual in comparison to how the other parties voted. The
pattern suggests manipulation (i.e., fraud).

FINAL President Attorney General Auditor General Treasurer

Democrat 3,458,289 3,465,215 3,128,811 3,239,093
Republican 3,377,674 3,153,677 3,337,860 3,291,742
Libertarian 79,380 120,478 205,905 152,892
Green 71,069 78,586 83,787
Total 6,915,343 6,810,439 6,751,162 6,767,514
As of 12/14/2020 10:48:52 AM EST

Figure 8: Voting consistency by Party


25. Inspection of the Mail-In Votes by party also shows the same uniform pattern in
the Green Party. This pattern also suggests manipulation (i.e., fraud).
Mail In Votes 2020 President Attorney General Auditor General Treasurer
Democrat 1,995,691 1,941,485 1,802,782 1,842,039
Republican 595,538 574,052 660,030 636,780
Libertarian 24,783 33,501 58,510 42,863
Green 30,491 33,834 38,034
TOTAL 2,616,012 2,579,529 2,555,156 2,559,716
As of 12/14/2020 10:48:52 AM EST

Figure 9: Mail-In Votes by Party for each State-wide Candidate

26. Given the results above, a reasonable person would expect that 30,000 Green
Party voters mailed in ballots and wrote-in the Presidential candidate, however
the official data shows only 2,541 did. Interestingly, Figure 10 (below) has been
removed from the Pennsylvania Secretary of State web-site.

Figure 10: Mail In Votes by Political Party

27. The low number (8,136) of Libertarian mail-in votes also appears improbable
given that they cast between 24K and 58K mail-in ballots.

28. Turning attention back to total votes among the state-wide offices, the Green
Party purportedly cast between 71K and 83K votes for the three state-wide office
contests. In 201611, while not running an AG candidate, Green Party candidates
received 159K votes for Auditor General and 170K for Treasurer (see Figure 11,
below). The decline from 2016 suggests that votes were shifted away from the
Green Party to another candidate.

President Attorney General Auditor General Treasurer

2020 Lib 79,380 120,478 205,905 152,892
2016 Lib 146,715 NA 235,142 132,654
2016 Green 49,941 NA 158,942 170,275
2020 Green NA 71,069 78,586 83,787

Figure 11: Comparing 2016 to 2020 Green and Libertarian votes

29. Libertarian votes in 2020 appear suspicious because the Presidential candidate
has less votes than the down ballot candidates. However, I compared the results
of state-wide elections in other states (Oregon, West Virginia, Montana, and
Missouri) and saw the same pattern; down ballot candidates outperformed the
presidential candidate, thus the Jorgenson 2020 vote total is correct.

30. The Pennsylvania Green Party and the Democrat Party engaged in a legal
dispute over the validity of signatures gathered for inclusion on the ballot,
alleging that the signatures were from dead people, wrong addresses, and
illegible signatures. After the Green Party provided evidence to disprove the
Democrats charges, the Democrats persisted with legal action to keep only the
Green Party Presidential and Vice-Presidential candidates off the ballot (citing


irregularities in paperwork).12 Write-in votes by the Green Party (and others)
would require adjudication by the Dominion Democracy Suite,13 ES&S EVS
6030.14 and ClearVote 2.015 systems used in Pennsylvania. 16

31. The various inconsistencies in Green Party voting patterns, the Greens writing
in their Presidential candidate (that will require system adjudication), and the
lack of transparency/detail17 of election data combined to create an environment
conducive to the switching of Green Party presidential votes (estimate 20K).

Inconsistencies: Major Party Voting Patterns

32. There are inconsistencies in the voting patterns of the major parties at the
state-wide level that cannot be attributed to chance and suggest vote
swapping.18 Examples include the inconsistent vote totals between Election
Day (in person) voting and Mail-In Voting between Democrats Biden and
Shapiro as well as Trump and the GOP down ballot candidates for Auditor
General and Treasurer (top vote getter(s) in green).

Election Day Votes President Attorney General Auditor General Treasurer

Democrat 1,409,341 1,461,226 1,273,298 1,342,729
Republican 2,731,230 2,530,207 2,626,111 2,603,048
Mail In Votes 2020 President Attorney General Auditor General Treasurer
Democrat 1,995,691 1,941,485 1,802,782 1,842,039
Republican 595,538 574,052 660,030 636,780
As of 12/14/2020 10:48:52 AM EST
For example, Libertarian and Green Party voters were included under “Minor Party” in publicly available reports
regarding mail-in ballots and are grouped under “Other” in official party registration reports.
See, page 18,

33. Based on voting patterns, it is highly unlikely that the GOP candidates for
Auditor General and Treasurer would get more mail-in votes than the President
considering that Trump got many more votes than they did on Election Day.

34. Similarly, it is also unlikely that the Democrat AG candidate would outpace
Biden by over 50,000 votes on Election Day and then lose to Biden by nearly the
same margin during mail-in voting. S. Stanley Young, PhD19 performed chi
square testing on these data for Allegheny County and commented that “this
discrepancy is not explainable by chance and it suggests the possibility of either
vote dumping to Biden and/or vote switching from Trump to Biden.”


The irregularities in voting patterns, scenario analysis of possible outcomes, and

analysis of available election results suggest that approximately 280,000 to 310,000
Pennsylvania voters had their votes changed in some manner (i.e., software, human
intervention, or introduction of malware) and that the magnitude of suspected
manipulation was/is great enough to change the outcome of the Pennsylvania



Scenario 1: STATUS QUO, assumes party loyalty.

Outside OFFICIAL Party Outside Outside
OFFICIAL Mail In Ballots Cast Loyalty Same Party Party MAIL IN Disloyal Votes Vote obtained
TOTAL VOTES By Party Factor Support Support VOTES PCT Remaining Available PCT
BIDEN 3,392,358 (D) 1,711,192 100% 1,711,192 257,200 1,968,392 0% 0 318,223 81%
TRUMP 3,338,021 ( R) 627,302 93.99% 589,573 0 589,573 6% 37,729 280,494 0%
JORGENSON 78,238 (L) 8,136 100% 8,136 16,360 24,496 0% 0 334,583 5%
Other O 296,854 0 23,294
Not Cast 23,294 8% 61,023
TOTAL 6,808,617 2,643,484 296,854 2,582,461 14%
Source: November 13, 2020

Under Scenario 1, Mail-in votes are held constant and 100% party loyalty. Biden
and Jorgenson receive 100% support from their bases, while Trump gets 94%.
Interestingly, under that support scenario Trump gets no outside party votes while
Biden gets 257,200 votes from outside his party. All told Biden gets an unrealistic
81% of support from all sources outside the Democrat Party.

Scenario 2: TOTAL MAIL IN VOTES ARE FIXED; no freedom of movement

Base not 100% loyal; polling factor applied.
Outside OFFICIAL Outside Outside
Mail In Ballots Cast Polling Same Party Party MAIL IN Disloyal Disloyal Vote obtained
TOTAL VOTES By Party Loyalty Support Support VOTES PCT Remain Available PCT
BIDEN 3,392,358 (D) 1,711,192 92% 1,574,297 394,095 1,968,392 8% 136,895 318,223 124%
TRUMP 3,338,021 ( R) 627,302 91% 570,845 18,728 589,573 9% 56,457 433,749 4%
JORGENSON 78,238 (L) 8,136 100% 8,136 16,360 24,496 0% 0 490,207 3%
Other O 296,854 0 0
Missing Mail In
TOTAL 6,808,617 2,643,484 429,184 2,582,461 0 193,353
Biden -136,895
Trump -18,728

Under Scenario 2, Mail In votes remain fixed. Party loyalty is adjusted based on
CNN exit polling data (except for Jorgenson) such that Democrats are 92% loyal
while Republicans are 91% loyal. At the same time, CNN’s cross-over voting shows
8 percentage points (out of 9 possible) for Biden and 7 percentage points (out of 8
possible) for Trump. In order for Biden to reach the fixed mail in total of 1.96M, he
has to obtain more votes than are available (124%).

Scenario 3: TOTAL MAIL IN VOTES and TOTAL VOTES have freedom of movement
Polling factors applied to Base, Other, and Disloyals
Outside REVISED Disloyal Outside Outside
REVISED Mail In Ballots Cast Polling Same Party Party MAIL IN Disloyal Party Vote obtained
TOTAL VOTES By Party Loyalty Support Support VOTES PCT Votes Available PCT
BIDEN 3,202,811 (D) 1,711,192 92% 1,574,297 204,548 1,778,845 8% 136,895 336,951 61%
TRUMP 3,553,332 ( R) 627,302 91% 570,845 234,039 804,884 9% 56,457 417,389 56%
JORGENSON 78,238 (L) 8,136 100% 8,136 16,360 24,496 0% 0 490,207 3%
Other O 296,854 0 0
Missing Mail In [35,259]
TOTAL 6,834,381 2,643,484 2,608,225 0 193,353 Deltas
Trump 44% of Other 87.50% of Disloyal Democrats Biden -189,547
Biden 52% of Other 88.89% of Disloyal Republicans Trump 215,311

computed from the mail-in vote statistics and polling data. Party loyalty and cross-
over vote percentages are the same as in Scenario 2; however, Other votes are split
between Biden (52%) and Trump (44%). As the results show, Trump increases his
mail-in vote total by 215K votes while Biden loses 189K votes. Also notable is that
Trump and Biden’s percent of Outside votes (last column) reflects a realistic 61% of
available votes obtained by Biden – contrasted with 124% in Scenario 2. Trump’s
percent of Outside votes also reflects a realistic 56% -- contrasted with 4% in
Scenario 2.

In summary, the Scenario analysis supports the claim that 220,000 votes were
removed from Trump in the 2020 Pennsylvania general election.