Third straw poll to choose the next UN Secretary-General

Discouraging votes on the rise

By Javier Surasky

Date drafted: September 18, 2026

Versión en español (ES)

Introduction

On September 18, the Security Council held its third straw poll as part of the process to select the person who will succeed António Guterres as Secretary-General of the United Nations.

In this article, I once again apply the model used to analyze the first straw poll, now incorporating the results obtained from the two subsequent rounds. Those interested in the details of the calculation should therefore refer to the first installment in this series tracking the straw polls.

Recall that, for the Security Council to formally recommend a candidate, it needs at least nine votes in favor, with no negative vote from any of its five permanent members. Since ballots distinguishing votes cast by permanent members from those cast by non-permanent members have not yet been used, the model can only measure the combinatorial uncertainty created by this lack of information.

Results of the third straw poll

Candidate	Encourage	Discourage	No opinion
Rebeca Grynspan	9	5	1
Carolyn Rodrigues-Birkett	8	6	1
Olara Otunnu	5	7	3
Rafael Mariano Grossi	5	8	2
Macky Sall	5	9	1
María Fernanda Espinosa	3	6	6
Michelle Bachelet	1	11	3
Ivonne A-Baki	0	10	5

Source: 1 for 8 billion.

Grynspan is the only candidate to reach the required nine votes, but Rodrigues-Birkett is only one vote short of that threshold. Otunnu, Grossi, and Sall form a second group, each with five votes of support, although their different numbers of discourage votes affect their prospects for the position.

The main takeaway from this round is not the candidates’ rankings but the sharp drop in no-opinion votes from the previous round, alongside a broad increase in discourage votes.

From the second to the third straw poll

Candidate	E2	E3	ΔE	D2	D3	ΔD	N2	N3	ΔN
Rebeca Grynspan	7	9	+2	4	5	+1	4	1	−3
Carolyn Rodrigues-Birkett	8	8	0	3	6	+3	4	1	−3
Olara Otunnu	5	5	0	5	7	+2	5	3	−2
Rafael Mariano Grossi	7	5	−2	6	8	+2	2	2	0
Macky Sall	6	5	−1	7	9	+2	2	1	−1
María Fernanda Espinosa	4	3	−1	3	6	+3	8	6	−2
Michelle Bachelet	2	1	−1	6	11	+5	7	3	−4
Ivonne A-Baki	0	0	0	8	10	+2	7	5	−2

Source: author’s own calculations.

Because the votes are secret, these variations do not allow us to say that a particular State changed its vote or to reconstruct individual shifts. They only provide evidence of aggregate net differences between the two rounds.

Across all eight candidates, the second round produced 39 encourage votes, 42 discourage votes, and 39 no-opinion votes. In the third round, those figures changed to 36, 62, and 22, respectively. This means three fewer votes of support, 20 more discourage votes, and 17 fewer no-opinion votes, suggesting that support and no-position votes from previous rounds are shifting toward rejection.

These totals describe 120 separate decisions—fifteen members assessing eight candidates—not the distribution of a single pool of votes among candidates. Each Council member may encourage or discourage more than one candidate.

The trend also appears when comparing the seven candidates who took part in all three rounds:

Aggregate vote type	First round	Second round	Third round
Encourage	45	39	36
Discourage	24	34	52
No opinion	36	32	17

Source: author’s own calculations.

Note: Ivonne A-Baki was not a candidate when the first round took place.

The process has become more clearly defined than it was at the outset, mainly because of the net increase in opposition. The data are consistent with a strategic hardening of Security Council members’ positions, already suggested in the second straw poll, and/or with the candidates’ inability to consolidate support. The aggregate results do not allow us to determine which interpretation is correct.

Distance from the threshold and P5 uncertainty

Candidate	E	D	N	Deficit	Deficit/N	Minimum D votes to reverse	No P5 D	At least one P5 D
Rebeca Grynspan	9	5	1	0	—	0	8.4%	91.6%
Carolyn Rodrigues-Birkett	8	6	1	1	100%	0	4.2%	95.8%
Olara Otunnu	5	7	3	4	133.3%	1	1.9%	98.1%
Rafael Mariano Grossi	5	8	2	4	200%	2	0.7%	99.3%
Macky Sall	5	9	1	4	400%	3	0.2%	99.8%
María Fernanda Espinosa	3	6	6	6	100%	0	4.2%	95.8%
Michelle Bachelet	1	11	3	8	266.7%	5	0%	100%
Ivonne A-Baki	0	10	5	9	180%	4	0.03%	99.97%

Source: author’s own calculations.

The “Deficit/N” column shows what proportion of the available no-opinion votes would have to turn into support to close the gap to nine. Values above 100% indicate that those votes are not enough and that the candidate also needs to reverse one or more discourage votes.

The final two columns show a combinatorial distribution based on an assumption of indifference: without information about who cast the negative votes, all possible combinations are treated symmetrically and are not intended to describe the political behavior of the permanent members.

Reading the candidacies

Rebeca Grynspan

• Grynspan regains first place in favorable votes and once again reaches the nine-vote threshold after her decline in the second round, but she continues to show an important weakness: her support went from 10 to 7 and then back to 9, while her discourage votes increased from 1 to 4 and now stand at 5. The recovery in favorable votes coexists with greater opposition than in either of the two previous rounds. With five discourage votes, the combinatorial probability that none of them comes from a permanent member is just 8.4%. This means that, as long as the origin of the negative votes remains unknown, the favorable result of nine votes of support remains surrounded by considerable uncertainty about whether one of the P5 is“blocked” here.

Carolyn Rodrigues-Birkett

• Rodrigues-Birkett retains her eight votes of support, but moves from three to six discourage votes and from four no-opinion votes to one. One vote short of the required threshold, her overall position becomes less favorable: she needs her only no-opinion vote to turn into support, and even then it would still be necessary to know whether any of her six discourage votes come from a P5 member. The combinatorial probability that at least one does is 95.8%.

Olara Otunnu

• Otunnu retains the five votes of support he reached in the second round, but adds two discourage votes and loses two no-opinion votes. With that, the strong progress he showed between the first and second ballots stalls: his three no-opinion votes are now insufficient to close a four-vote deficit and reach the required nine, meaning he needs to reverse discourage votes.

Rafael Mariano Grossi

• The warning that emerged from the second straw poll becomes clearer: Grossi faces strong opposition and also loses two votes of support, moving from 7-6-2 to 5-8-2, which means that he too needs to reverse discourage votes. His position is no longer that of a candidate close to the threshold, and he now faces substantial arithmetic and political obstacles.

Macky Sall

• Sall loses one vote of support, adds two discourage votes, and is left with a single no-opinion vote, joining the broad group of candidates who need to reverse negative votes. In combinatorial terms, this is one of the more fragile situations.

María Fernanda Espinosa

• Espinosa continues on an adverse trajectory: her support falls from five votes in the first round to four in the second and three in the third, while her discourage votes rise from two to three and then to six. Even so, her six no-opinion votes are arithmetically sufficient to close the support deficit, but only if she manages to convert all of them into support. The calculation shows a real possibility, but a politically complex one.

Michelle Bachelet

• Bachelet records the sharpest deterioration of the entire process: her support falls from six to two and then to one, while discourage votes rise from five to six and finally to eleven. Since the Council has ten non-permanent members, at least one of those eleven discourage votes must come from a permanent member. Her selection is essentially ruled out, although possible changes in votes mean that we must continue to consider her.

Ivonne A-Baki

• A-Baki again receives no support, while discourage votes rise from eight in the previous straw poll to ten in this one, with no-opinion votes falling to five. This means she would also need to reverse discourage votes, and the combinatorial probability that at least one of her ten negative votes comes from a permanent member is 99.97%.

The questions from the second analysis

In the article on the second round, I identified five elements worth watching in the third straw poll. The results provide some indications on those points:

1. Grynspan reverses her decline in favorable votes, although the number of discourage votes continues to increase.
2. Rodrigues-Birkett does not secure the ninth vote of support and doubles her negative votes.
3. Grossi does not reduce opposition and loses two votes of support, clearly weakening his candidacy.
4. A-Baki remains at zero favorable votes and adds new discourage votes.
5. No differentiated identification of votes cast by permanent and non-permanent members has been released: P5 uncertainty therefore remains central except in Bachelet’s case, where the existence of at least one discourage vote from a permanent member is arithmetically certain.

The third round also provides information on no-opinion votes. Across the eight candidates, these decrease by seventeen, discourage votes increase by twenty, and support falls by three.

That balance provides less support for the hypothesis that no-opinion votes mainly represented a reserve of future support and is more consistent with the idea that positions on the existing candidates are becoming more clearly defined in a negative direction. Does this open the door to a “dark horse”?

What to watch in the next round

The fourth round should make it possible to assess five main questions:

1. Whether Grynspan can remain at nine or more votes of support while also reducing her discourage votes.
2. Whether Rodrigues-Birkett secures the additional vote she needs or continues to accumulate opposition.
3. Whether any of the three candidates with five votes of support manages to break away from the group, or whether that group begins to lose viability.
4. Whether the candidates in the weakest positions reconsider remaining in the race.
5. Whether differentiated ballots finally appear, making it possible to determine how many discourage votes come from permanent members.

Conclusions

Grynspan is the only candidate to reach the required nine votes, placing her in the strongest position, although her five discourage votes prevent this round from being read as a consolidation of her lead.

Rodrigues-Birkett remains the closest alternative, but she does not increase her support and faces growing opposition. Meanwhile, the other candidates are moving farther away from the required threshold, and only Espinosa could reach it without having to reverse votes that are currently negative. But doing so would require every one of her no-opinion votes to turn into support, without exception.

The most important signal from the third straw poll is collective: the previously mentioned shift in no-opinion votes, which are falling quickly and moving toward discourage votes.

The Council appears to be leaving the initial evaluation stage behind. What we do not know is whether it is moving closer to a consensus or making visible the absence of a candidate capable of bringing one together.