ANALYSIS: Meloni Holds 27%, but Vannacci Is Rearranging Italy’s Right

Introduction

Fratelli d’Italia stood at 27% in the SWG poll broadcast on August 3, 2026 by Tg La7, unchanged from July 27. That is stability in one measurement, not a declaration that Italy’s right has stopped changing.

BiDiMedia, published on August 6, 2026, puts confidence in Giorgia Meloni at 35%, down one point from mid-July, and distrust at 40%. Its enlarged centre-right coalition including Roberto Vannacci’s Futuro Nazionale reaches 47.2%.

The two polls do not use the same method, reference dates or disclosed uncertainty. SWG gives a margin of error of plus or minus 2.8 percentage points; the assigned record does not provide an explicit margin for BiDiMedia. The comparison needs care before it can carry a political verdict.

SWG’s 27% is stable against July 27

Fratelli d’Italia is placed at 27%

The public record starts with Fratelli d’Italia and 27%. The SWG poll for Tg La7 places Fratelli d’Italia at 27% of voting intentions on August 3. The file places August 3, 2026 in the documented account under “Fratelli d’Italia is placed at 27%”.

Its practical consequence is narrow but real. The result is a snapshot of stated intention under SWG’s method, not an election result. The supplied material ties that reading to SWG, keeps the voting-intention result in view, and does not establish an election outcome beyond the stated evidence. The mandate has begun.

A stable vote share can coexist with a shifting coalition.

The comparison point is July 27

On the documented timeline, July 27, 2026 and stable. SWG reports the 27% figure as stable compared with July 27. The file places 27% in the documented account under “The comparison point is July 27”.

That matters because the wording fixes the boundary of the claim. Stability is meaningful only within that stated comparison and its margin of error. The supplied material ties that reading to the SWG comparison, keeps the stated result in view, and does not establish absolute permanence beyond the stated evidence. The date fixes the claim.

A margin of error is part of a poll result, not fine print.

The PD slips to 20.6%

Partito Democratico is at 20.6%

The stated mandate places Partito Democratico and 20.6%. The SWG reading puts the Partito Democratico at 20.6%. The file places SWG in the documented account under “Partito Democratico is at 20.6%”.

The consequence is institutional, not theatrical. The number belongs to the same August 3 poll and should not be fused with BiDiMedia’s separate measures. The supplied material ties that reading to the August 3 poll, keeps the party result in view, and does not establish a cross-poll trend beyond the stated evidence. The wording carries weight.

A tenth of a point is movement, but not a mandate.

The reported change is minus 0.3 points

At the centre of this file, minus 0.3 points and July 27. SWG records the PD as down 0.3 points from its July 27 comparison. The file places plus or minus 2.8 points in the documented account under “The reported change is minus 0.3 points”.

This changes the frame without proving an outcome. A change that small sits inside the context of SWG’s stated plus-or-minus 2.8-point margin. The supplied material ties that reading to the reported change, keeps the margin in view, and does not establish a definitive shift beyond the stated evidence. The outcome is not documented.

Confidence and voting intention are different measurements.

M5S reaches 13% with a small gain

Movimento 5 Stelle is at 13%

The published account identifies Movimento 5 Stelle and 13%. SWG places the Movimento 5 Stelle at 13% in the August 3 broadcast. The file places August 3, 2026 in the documented account under “Movimento 5 Stelle is at 13%”.

The point is measurable even where the record is incomplete. The figure adds another element to the poll’s balance without resolving the coalition story on its own. The supplied material ties that reading to SWG, keeps the poll balance in view, and does not establish the coalition story beyond the stated evidence. The record stays narrow.

Forty percent distrust changes the picture without predicting an election.

The gain is 0.1 point

That same account links plus 0.1 point and SWG. The same poll records a gain of 0.1 point for the Movimento 5 Stelle. The file places plus or minus 2.8 points in the documented account under “The gain is 0.1 point”.

The available evidence supports a limited conclusion. The movement is reported, but it is smaller than the survey’s declared margin of error. The supplied material ties that reading to the reported movement, keeps the margin in view, and does not establish a decisive advance beyond the stated evidence. The evidence names the actor.

Vannacci's effect appears in a coalition calculation, not a direct comparison.

SWG’s margin of error frames every small change

The declared margin is plus or minus 2.8 points

A separate part of the record shows plus or minus 2.8 points and SWG. The SWG poll discloses a margin of error of plus or minus 2.8 percentage points. The file places margin of error in the documented account under “The declared margin is plus or minus 2.8 points”.

That is the pressure point in the file. That disclosure is part of the result and belongs beside every small week-to-week variation. The supplied material ties that reading to the disclosed uncertainty, keeps the result in view, and does not establish precision beyond the method beyond the stated evidence. The file gives no shortcut.

The expanded bloc gains little, but the internal arithmetic changes.

The method combines three channels

The available material also names CAWI and CATI and CAMI. The poll uses a mixed online, telephone and mobile method: CAWI, CATI and CAMI. The file places mixed method in the documented account under “The method combines three channels”.

The difference lies in what has actually been stated. Method tells readers how SWG collected the data; it does not make separate institutes directly interchangeable. The supplied material ties that reading to the SWG methodology, keeps the collection mode in view, and does not establish interchangeable polling beyond the stated evidence. The claim has an author.

Indecision makes every clean story less certain.

BiDiMedia measures confidence differently

Confidence in Meloni is listed at 35%

This is where the paper trail narrows: Giorgia Meloni and 35%. BiDiMedia, published on August 6, puts confidence in Giorgia Meloni at 35%. The file places BiDiMedia in the documented account under “Confidence in Meloni is listed at 35%”.

This is a constraint on interpretation, not a reason to inflate it. Confidence is not the same indicator as party voting intention, so the 35% should not be read as a Fratelli d’Italia vote share. The supplied material ties that reading to the confidence measure, keeps the indicator in view, and does not establish a party vote share beyond the stated evidence. The limit belongs in print.

Two pollsters do not create one interchangeable series.

The confidence figure is down one point

Another documented element is minus one point and mid-July. The BiDiMedia comparison says confidence in Meloni fell one point from mid-July. The file places BiDiMedia in the documented account under “The confidence figure is down one point”.

The record supplies an anchor and withholds the rest. The change is reported against BiDiMedia’s own reference point, not against SWG’s July 27 line. The supplied material ties that reading to the comparison, keeps the reference point in view, and does not establish a SWG comparison beyond the stated evidence. The source draws the line.

An editorial observation is not a polling methodology.

Distrust reaches 40% in BiDiMedia

BiDiMedia puts distrust at 40%

The record then turns to 40% and distrust. The assigned record says BiDiMedia measured distrust in Meloni at 40%. The file places BiDiMedia in the documented account under “BiDiMedia puts distrust at 40%”.

That leaves a specific test for the next public development. That measure broadens the picture of personal confidence but does not itself allocate parliamentary seats. The supplied material ties that reading to the stated measure, keeps the reading in view, and does not establish parliamentary seats beyond the stated evidence. The next proof must be public.

The date of comparison determines what a change means.

Confidence and distrust are separate indicators

One constraint appears in plain view: 35% confidence and 40% distrust. The two BiDiMedia figures appear together in the record as different measures of confidence and distrust. The file places BiDiMedia in the documented account under “Confidence and distrust are separate indicators”.

The wording creates a responsibility without describing its completion. They show public sentiment in that poll; they are not a replacement for vote-intention data. The supplied material ties that reading to the two indicators, keeps the sentiment reading in view, and does not establish vote intention beyond the stated evidence. The promise still needs evidence.

Meloni's 27 percent is a snapshot, not an exemption from pressure.

Vannacci changes the coalition arithmetic

Futuro Nazionale enters the expanded bloc

The next item is not a forecast but Futuro Nazionale and Roberto Vannacci. BiDiMedia’s enlarged centre-right coalition includes Roberto Vannacci’s Futuro Nazionale. The file places expanded coalition in the documented account under “Futuro Nazionale enters the expanded bloc”.

This is the part that resists an easy headline. The party’s relevance in this record is seen through coalition totals, not through a direct individual party score in the supplied block. The supplied material ties that reading to the BiDiMedia frame, keeps the party’s role in view, and does not establish an individual score beyond the stated evidence. The number has a boundary.

The traditional bloc's decline is stated within BiDiMedia's own comparison.

The enlarged centre-right reaches 47.2%

In the source material, 47.2% and plus 0.2 points. BiDiMedia puts the enlarged centre-right at 47.2%, up 0.2 points. The file places BiDiMedia in the documented account under “The enlarged centre-right reaches 47.2%”.

The source shows a direction, not a finished result. The gain is modest, but the comparison becomes more interesting when set beside the traditional bloc’s reported decline. The supplied material ties that reading to the enlarged coalition, keeps the reported gain in view, and does not establish a sweeping mandate beyond the stated evidence. The record does not guess.

Without a BiDiMedia margin, caution is not optional.

The traditional centre-right is at 40.1%

The coalition without Vannacci is at 40.1%

The factual hinge is 40.1% and traditional centre-right. BiDiMedia gives the traditional centre-right, without Vannacci, 40.1%. The file places without Vannacci in the documented account under “The coalition without Vannacci is at 40.1%”.

The distinction protects the record from a false shortcut. That figure is an internal coalition comparison in one poll, not a direct measurement of a different electorate. The supplied material ties that reading to the BiDiMedia comparison, keeps the coalition line in view, and does not establish a separate electorate beyond the stated evidence. The comparison has a date.

Italy's right can hold its lead while rearranging its own voters.

The traditional bloc is down 0.7 points

A second limit matters here: minus 0.7 points and BiDiMedia. The record says the traditional centre-right declined 0.7 points in BiDiMedia’s comparison. The file places traditional centre-right in the documented account under “The traditional bloc is down 0.7 points”.

That boundary is part of the story itself. The contrast with the enlarged bloc is evidence of a possible internal transfer, not proof of why individual voters moved. The supplied material ties that reading to the comparison, keeps the internal shift in view, and does not establish voter motives beyond the stated evidence. The public statement stands alone.

Undecided voters limit every clean reading

BiDiMedia reports 23% undecided

The stated sequence begins with 23% and BiDiMedia. The assigned material gives BiDiMedia an undecided share of 23%. The file places undecided voters in the documented account under “BiDiMedia reports 23% undecided”.

The consequence belongs to the institution named in the source. A share that large is a reminder that any coalition arithmetic remains contingent on voters who have not settled. The supplied material ties that reading to the polling limit, keeps the uncertainty in view, and does not establish a fixed electorate beyond the stated evidence. The paper trail is specific.

SWG reports 30% undecided

The documentary trail records 30% and SWG. The same material gives SWG an undecided share of 30%. The file places undecided voters in the documented account under “SWG reports 30% undecided”.

The documentation does not establish more than this. The higher figure belongs to a different pollster and method, reinforcing the case against a point-for-point merger. The supplied material ties that reading to the separate poll, keeps the difference in view, and does not establish a merged series beyond the stated evidence. The mechanism is now visible.

Different methods block a mechanical comparison

SWG and BiDiMedia use different designs

The file distinguishes SWG and BiDiMedia. The limitation states that the two polls use slightly different methodologies and field dates. The file places different methodologies in the documented account under “SWG and BiDiMedia use different designs”.

The public claim therefore remains tied to its author. That is enough to block a mechanical comparison of every change between the two institutes. The supplied material ties that reading to the stated limitation, keeps the two surveys in view, and does not establish point-by-point equivalence beyond the stated evidence. The claim cannot outrun its source.

Only SWG’s margin is explicit

The reported figure concerns plus or minus 2.8 points and BiDiMedia. The fact block explicitly provides SWG’s margin of error but not a corresponding BiDiMedia margin. The file places explicit margin in the documented account under “Only SWG’s margin is explicit”.

The figure has meaning only with its date and attribution. The missing disclosure does not invalidate BiDiMedia; it sets a limit on comparative confidence in this article. The supplied material ties that reading to the disclosure gap, keeps the limitation in view, and does not establish matched uncertainty beyond the stated evidence. The distinction matters.

The published commentary keeps its status

The Italian quote is an editorial observation

No broader conclusion is needed to see Cala la fiducia in Meloni and Blasting News. Blasting News described confidence in Meloni as falling from the previous July period. The file places BiDiMedia in the documented account under “The Italian quote is an editorial observation”.

The evidence is stronger on the action than on its effects. That wording is editorial commentary on BiDiMedia’s result, not the methodology of the institute itself. The supplied material ties that reading to the editorial observation, keeps the quote in view, and does not establish a methodology statement beyond the stated evidence. The sequence is documented.

The polling data must remain the base

The material supplied for publication says BiDiMedia and SWG. The article uses the stated numbers from BiDiMedia and SWG while keeping the media description separate from the raw polling frame. The file places polling data in the documented account under “The polling data must remain the base”.

That is why the qualification cannot be discarded. That distinction protects the analysis from turning a commentary line into a measured fact. The supplied material ties that reading to the analysis, keeps the stated numbers in view, and does not establish a media verdict beyond the stated evidence. The uncertainty remains named.

Meloni’s stability is real but bounded

The 27% belongs to a dated snapshot

A careful reading keeps 27% and August 3, 2026. Fratelli d’Italia’s 27% is a dated SWG result, stable against its July 27 comparison. The file places July 27 in the documented account under “The 27% belongs to a dated snapshot”.

The material separates a formal statement from a demonstrated result. It establishes stability in that reading; it does not cancel the separate confidence and coalition movement recorded elsewhere. The supplied material ties that reading to the SWG snapshot, keeps the stability in view, and does not establish all other movement beyond the stated evidence. The evidence stops there.

The right can shift within its own lead

This portion of the evidence establishes 47.2% and 40.1%. BiDiMedia’s enlarged 47.2% bloc and traditional 40.1% bloc place Vannacci inside the right’s internal arithmetic. The file places Vannacci in the documented account under “The right can shift within its own lead”.

This does not settle the larger dispute, but it fixes one point. The figures suggest recomposition within the coalition, with the source’s methodological limits still attached. The supplied material ties that reading to the coalition arithmetic, keeps the recomposition in view, and does not establish a final electoral verdict beyond the stated evidence. The consequence is concrete.

The next data point needs the same caution

Dates and bases decide the meaning of change

The source does not turn this into certainty; it does show August 3, 2026 and August 6, 2026. The reports were published on August 3 and August 6, with different reference comparisons. The file places reference comparisons in the documented account under “Dates and bases decide the meaning of change”.

The documented limit is as important as the reported fact. The dates make clear why apparent movement cannot be reduced to one unbroken line. The supplied material ties that reading to the publication dates, keeps the separation in view, and does not establish one unbroken series beyond the stated evidence. The record retains its limits.

The evidence supports a qualified conclusion

That distinction is carried by 27% and 35%. The record supports a stable 27% party reading in SWG and a 35% confidence reading in BiDiMedia. The file places SWG and BiDiMedia in the documented account under “The evidence supports a qualified conclusion”.

The result is a record that demands precision rather than volume. It supports a changing picture, not certainty about the direction or durability of every individual voter. The supplied material ties that reading to the qualified conclusion, keeps the two readings in view, and does not establish voter certainty beyond the stated evidence. The file refuses a shortcut.

Conclusion

The central Italian fact is not that Giorgia Meloni has either won or lost an argument in a single week. It is that 27% in SWG can coexist with 35% confidence in BiDiMedia and an enlarged right-wing coalition at 47.2%.

That coexistence is not a contradiction once the measures, methods, dates and margins are kept separate. It is a warning against the easy habit of treating every poll as the same instrument.

The lead holds. The coalition is moving.

Signature

Signed Maxime Marquette, columnist

Columnist's Transparency box

Editorial positioning

This analysis treats Italian polling as evidence with limits, not as a device for declaring an election before it happens. It makes no claim about individual voter motives beyond the data supplied.

The article distinguishes party intention, personal confidence, coalition totals and undecided shares.

Methodology and sources

The factual base is limited to the assigned fact block and its stated source trail. Dates, figures and named institutions are retained with their attribution.

No information from outside that supplied record is used to fill gaps in the chronology, the evidence or the consequences.

Nature of the analysis

The piece is a documented analysis of the public material rather than a finding of fact beyond the published record.

Claims whose status is disputed, preliminary or source-bound remain described that way throughout the article.

Sources

Primary sources

The assigned material identifies SWG and BiDiMedia as the polling institutes, with results published through the cited reports.

The publication record below keeps the supplied source URLs in the final section only.

Secondary sources

These items are the reports and source pages named in the assigned fact block.

Repeated links, where present, preserve the source set used by the associated edition.

• QuiFinanza on the SWG poll — August 5, 2026
• Blasting News on the BiDiMedia poll — August 6, 2026
• Il Foglio on Vannacci polling — August 4, 2026
• Adnkronos polling report — 2026
• Reuters on Vannacci in opinion polls — July 24, 2026
• Quotidiano Nazionale polling report — 2026