Ricardo Ciacci Strina
@riccardociacci.bsky.social
Applied Micro-econometrician.
Associate Professor of Economics: @UComillas
Fields: Gender, Crime, Development.
@uc3m , @Columbia & @EUI-EU alumnus
#HeForShe
Associate Professor of Economics: @UComillas
Fields: Gender, Crime, Development.
@uc3m , @Columbia & @EUI-EU alumnus
#HeForShe
y concluyendo erróneamente que rechazar la hipótesis nula es equivalente a aceptarla. La revista abrió una investigación y tras 10 meses de investigación un comité científico independiente estableció que: no había mala conducta por mi parte, 6/N
July 30, 2025 at 9:28 PM
y concluyendo erróneamente que rechazar la hipótesis nula es equivalente a aceptarla. La revista abrió una investigación y tras 10 meses de investigación un comité científico independiente estableció que: no había mala conducta por mi parte, 6/N
de acoso. Tuits del estilo “no entiendo como tu artículo pudo pasar la revisión por pares” o “no tienes datos para realizar una investigación científica” fueron escritos y algunos posteriormente borrados. Finalmente usaron canales más aptos para la comunicación científica 4/N
July 30, 2025 at 9:28 PM
de acoso. Tuits del estilo “no entiendo como tu artículo pudo pasar la revisión por pares” o “no tienes datos para realizar una investigación científica” fueron escritos y algunos posteriormente borrados. Finalmente usaron canales más aptos para la comunicación científica 4/N
—not only misrepresent my work but also set a dangerous precedent for academic integrity. What makes this situation particularly egregious is that, after a rigorous 10-month investigation conducted by an independent scientific committee, it was conclusively 2/N
July 30, 2025 at 12:03 PM
—not only misrepresent my work but also set a dangerous precedent for academic integrity. What makes this situation particularly egregious is that, after a rigorous 10-month investigation conducted by an independent scientific committee, it was conclusively 2/N
2) A difference-in-discontinuity approach, subtracting pre-treatment December–January placebo estimates from the above estimates (same approach as in Ciacci (2025))
In this case, results are robust for the polynomial of order two. 5/N
In this case, results are robust for the polynomial of order two. 5/N
July 21, 2025 at 12:04 PM
2) A difference-in-discontinuity approach, subtracting pre-treatment December–January placebo estimates from the above estimates (same approach as in Ciacci (2025))
In this case, results are robust for the polynomial of order two. 5/N
In this case, results are robust for the polynomial of order two. 5/N
To address potential multicollinearity concerns in Ciacci (2024), I conduct two additional tests:
1) Replicating Ciacci’s (2025) methodology to estimate a regionally representative coefficient (as in Ciacci 2024) 4/N
1) Replicating Ciacci’s (2025) methodology to estimate a regionally representative coefficient (as in Ciacci 2024) 4/N
July 21, 2025 at 12:04 PM
To address potential multicollinearity concerns in Ciacci (2024), I conduct two additional tests:
1) Replicating Ciacci’s (2025) methodology to estimate a regionally representative coefficient (as in Ciacci 2024) 4/N
1) Replicating Ciacci’s (2025) methodology to estimate a regionally representative coefficient (as in Ciacci 2024) 4/N
The findings also hold under the stringent robustness tests in Ciacci (2025), where the sample is expanded by adding one month around the cutoff until full coverage is achieved. Results remain stable for the period where pre- and post-treatment months are balanced 3/N
July 21, 2025 at 12:04 PM
The findings also hold under the stringent robustness tests in Ciacci (2025), where the sample is expanded by adding one month around the cutoff until full coverage is achieved. Results remain stable for the period where pre- and post-treatment months are balanced 3/N
1) Using a first-order polynomial
2) Removing seasonal fixed effects
3) Removing police as a control variable 2/N
2) Removing seasonal fixed effects
3) Removing police as a control variable 2/N
July 21, 2025 at 12:04 PM
1) Using a first-order polynomial
2) Removing seasonal fixed effects
3) Removing police as a control variable 2/N
2) Removing seasonal fixed effects
3) Removing police as a control variable 2/N
The thread below further illustrates the original decision made by the scientific committee — a decision that was only overturned following a social media harassment campaign, setting a troubling precedent that such campaigns can pressure 1/2
July 18, 2025 at 1:16 PM
The thread below further illustrates the original decision made by the scientific committee — a decision that was only overturned following a social media harassment campaign, setting a troubling precedent that such campaigns can pressure 1/2
This was never about honest academic inquiry. It was about senior professors pressuring a journal to retract a junior researcher’s paper simply because they did not like the findings 3/3
July 17, 2025 at 6:26 AM
This was never about honest academic inquiry. It was about senior professors pressuring a journal to retract a junior researcher’s paper simply because they did not like the findings 3/3
They sent me an email, and I provided the code within 28 hours. Their initial request even specified a tight deadline 'relatively quick' which I found out meant around 5 to 6 hours. Beyond that, I offered to collaborate on a paper to further the research, but that was also declined.
2/3
2/3
July 17, 2025 at 6:26 AM
They sent me an email, and I provided the code within 28 hours. Their initial request even specified a tight deadline 'relatively quick' which I found out meant around 5 to 6 hours. Beyond that, I offered to collaborate on a paper to further the research, but that was also declined.
2/3
2/3
The retraction notice relied on false claims—neither Zimmermann (2025) nor Ciacci (2025) state that the findings of Ciacci (2024) are incorrect. 12/N
July 11, 2025 at 11:23 AM
The retraction notice relied on false claims—neither Zimmermann (2025) nor Ciacci (2025) state that the findings of Ciacci (2024) are incorrect. 12/N
Recently, I discovered that the @SpringerNature decision to retract the article was not based on scientific evidence but was influenced by the harassment campaign. x.com/johannarickn... 11/N
July 11, 2025 at 11:23 AM
Recently, I discovered that the @SpringerNature decision to retract the article was not based on scientific evidence but was influenced by the harassment campaign. x.com/johannarickn... 11/N
I requested that if the retraction were to proceed, a note from me should be included to explain my position. That note was omitted. 10/N
July 11, 2025 at 11:23 AM
I requested that if the retraction were to proceed, a note from me should be included to explain my position. That note was omitted. 10/N
In late May, a member of Springer’s Research Integrity team asked me to misrepresent the findings of my own paper. Transparency, supposedly a core value of @springernature.com ’s Research Integrity team, was glaringly absent. 9/N
July 11, 2025 at 11:23 AM
In late May, a member of Springer’s Research Integrity team asked me to misrepresent the findings of my own paper. Transparency, supposedly a core value of @springernature.com ’s Research Integrity team, was glaringly absent. 9/N
I was instructed to remove the word Corrigendum. The article would instead be published under the “Additional Evidence” format. Once both papers were published, the harassment campaign persisted. 8/N
July 11, 2025 at 11:23 AM
I was instructed to remove the word Corrigendum. The article would instead be published under the “Additional Evidence” format. Once both papers were published, the harassment campaign persisted. 8/N
It was also determined that the two articles would be read in tandem (the editor explicitly confirmed this in response to my observation that the new article addressed a different research question and estimated a different target parameter). 7/N
July 11, 2025 at 11:23 AM
It was also determined that the two articles would be read in tandem (the editor explicitly confirmed this in response to my observation that the new article addressed a different research question and estimated a different target parameter). 7/N
The journal launched an investigation and concluded that no misconduct had occurred. However, I was asked to verify whether the results were influenced by multicollinearity. Ultimately, it was agreed that a Corrigendum would suffice to avoid retraction. 6/N
July 11, 2025 at 11:23 AM
The journal launched an investigation and concluded that no misconduct had occurred. However, I was asked to verify whether the results were influenced by multicollinearity. Ultimately, it was agreed that a Corrigendum would suffice to avoid retraction. 6/N
Instead, the same authors submitted a report to the journal requesting the retraction of my article, claiming that the observed effect could stem from a multicollinearity issue in my Stata code. 5/N
July 11, 2025 at 11:23 AM
Instead, the same authors submitted a report to the journal requesting the retraction of my article, claiming that the observed effect could stem from a multicollinearity issue in my Stata code. 5/N
either I responded within 5–6 hours, or he would post something on Twitter. Exemplary academic conduct, indeed. 3/N
July 11, 2025 at 11:23 AM
either I responded within 5–6 hours, or he would post something on Twitter. Exemplary academic conduct, indeed. 3/N
My reanalysis confirms that the large increase in reported rapes still appears, even under a completely different statistical approach.
The issue under review—multicollinearity—was not driving the original result. 2/4
The issue under review—multicollinearity—was not driving the original result. 2/4
July 3, 2025 at 3:11 PM
My reanalysis confirms that the large increase in reported rapes still appears, even under a completely different statistical approach.
The issue under review—multicollinearity—was not driving the original result. 2/4
The issue under review—multicollinearity—was not driving the original result. 2/4
Ciacci (2025) clearly finds that the results of Ciacci (2024) were not driven by multicollinearity. The specification most comparable to Ciacci (2024) shows an increase of approximately 56% in rape cases—robust even near the optimal bandwidth. Restoring the truth is paramount. @springernature.com
July 3, 2025 at 7:10 AM
Ciacci (2025) clearly finds that the results of Ciacci (2024) were not driven by multicollinearity. The specification most comparable to Ciacci (2024) shows an increase of approximately 56% in rape cases—robust even near the optimal bandwidth. Restoring the truth is paramount. @springernature.com
In the current society work serves both as a means of survival and as a search for meaning in life — a dual role insightfully explored by David Graeber @davidgraeberinst.bsky.social
July 2, 2025 at 11:05 PM
In the current society work serves both as a means of survival and as a search for meaning in life — a dual role insightfully explored by David Graeber @davidgraeberinst.bsky.social
This estimate remains stable near the optimal bandwidth. However, it does not hold up under the 70/80 regressions that I perform for each specification as robustness checks.
July 2, 2025 at 10:37 PM
This estimate remains stable near the optimal bandwidth. However, it does not hold up under the 70/80 regressions that I perform for each specification as robustness checks.
The additional evidence paper investigates a different parameter and research question: it focuses on nationally representative estimates using population weights. However, the specification most similar to Ciacci (2024)—unweighted, with seasonal FE and officer controls—finds the same result.
July 2, 2025 at 10:37 PM
The additional evidence paper investigates a different parameter and research question: it focuses on nationally representative estimates using population weights. However, the specification most similar to Ciacci (2024)—unweighted, with seasonal FE and officer controls—finds the same result.
I am deeply disheartened by the misrepresentation of my case in this still ongoing social media harassment campaign. The retraction falsely claims that the editor’s comment and my reanalysis declared Ciacci (2024) ‘incorrect’—a distortion of the record. 1/2
July 2, 2025 at 4:12 PM
I am deeply disheartened by the misrepresentation of my case in this still ongoing social media harassment campaign. The retraction falsely claims that the editor’s comment and my reanalysis declared Ciacci (2024) ‘incorrect’—a distortion of the record. 1/2