election

Tackling Cyber Disinformation in Elections: Applying International Human Rights Law

Research Event

6 November 2019 - 5:30pm to 7:00pm

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

Susie Alegre, Barrister and Associate Tenant, Doughty Street Chambers
Evelyn Aswad, Professor of Law and the Herman G. Kaiser Chair in International Law, University of Oklahoma
Barbora Bukovská, Senior Director for Law and Policy, Article 19
Kate Jones, Director, Diplomatic Studies Programme, University of Oxford
Chair: Harriet Moynihan, Associate Fellow, International Law Programme, Chatham House

Cyber operations are increasingly used by political parties, their supporters and foreign states to influence electorates – from algorithms promoting specific messages to micro-targeting based on personal data and the creation of filter bubbles.
 
The risks of digital tools spreading disinformation and polarizing debate, as opposed to deepening democratic engagement, have been highlighted by concerns over cyber interference in the UK’s Brexit referendum, the 2016 US presidential elections and in Ukraine. 
 
While some governments are adopting legislation in an attempt to address some of these issues, for example Germany’s ‘NetzDG’ law and France’s ‘Law against the manipulation of information’, other countries have proposed an independent regulator as in the case of the UK’s Online Harms white paper. Meanwhile, the digital platforms, as the curators of content, are under increasing pressure to take their own measures to address data mining and manipulation in the context of elections. 

How do international human rights standards, for example on freedom of thought, expression and privacy, guide the use of digital technology in the electoral context? What practical steps can governments and technology actors take to ensure policies, laws and practices are in line with these fundamental standards? And with a general election looming in the UK, will these steps come soon enough?
 
This event brings together a wide range of stakeholders including civil society, the tech sector, legal experts and government, coincides with the publication of a Chatham House research paper on disinformation, elections and the human rights framework

Jacqueline Rowe

Programme Assistant, International Law Programme
020 7389 3287




election

Iran Crisis Pushes Foreign Policy to Top of 2020 Election Debate

14 January 2020

Dr Lindsay Newman

Senior Research Fellow, US and the Americas Programme
Democrats would be wise to communicate a clear alternative to Trump’s ‘America First’ policy in the Middle East.

2020-01-14-Trump.jpg

Donald Trump speaks to the media in front of the White House on Monday. Photo: Getty Images.

Conventional wisdom says that foreign policy takes a backseat role in US elections. But last autumn’s Democratic primary debates suggest a potential shift is taking place in the conventional view. While healthcare dominated the discussion (Democrats attribute their 2018 midterm gains to the issue), through November foreign policy followed closely behind in second place in terms of minutes devoted to the discussion.

This trend is consistent with President Donald Trump’s America First approach to foreign policy, in which an eye is always kept on how decisions abroad play for the domestic audience. One former Trump administration official has called this dynamic the ‘recoupling’ of foreign policy with domestic policy.

The US–China trade conflict, which commanded headlines throughout 2019, is perhaps the best example of this recoupling, tying trade imbalances less with the geopolitical than with domestic impact on farmers. Immigration is another policy area in which Trump has linked domestic implications and indeed domestic opinion with foreign policy. It’s in the title: America First.

Now, for better or worse, the targeted killing of Qassem Soleimani, Iran’s response and the subsequent fallout may make US foreign policy towards Iran and the US role in the Middle East a central issue for the 2020 US elections. As it comes just ahead of the Democratic presidential primaries, voters will be looking to the candidates to differentiate their foreign policy experience and proposals for America’s Middle East policy.

To President Donald Trump, Soleimani’s assassination represents a campaign promise kept to confront Iran’s aggression.

The Trump administration initially justified the action by citing intelligence of an imminent threat to US personnel and targets, but after Defense Secretary Mark Esper called this into question, Trump tweeted that ‘it doesn’t really matter because of [Soleimani’s] horrible past’. Ultimately, Trump’s message, on the campaign trail and any general debate stage he agrees to be on, is that he has overseen a new national security strategy for Iran.

Soleimani’s removal from the Iranian calculus is just a part of this broader policy, which also includes neutralizing the Iranian government’s destabilizing influence in the Middle East, denying Iran and especially the Islamic Revolutionary Guard Corps’ access to funding for its malign activities, and rallying the international community against domestic human rights violations and unjust detentions.

To counter Trump, Democrats and democratic presidential candidates would be best-served by offering a simple argument that too links domestic interests and foreign policy: the killing of Soleimani and Trump’s national security strategy for Iran have not made the US or its interests safer.

Iran’s ballistic missile attack on US forces in Iraq, which Iran’s Supreme Leader Ayatollah Ali Khamenei called a ‘slap in the face’ for the US, makes the risks to US assets and personnel abundantly clear. Even if Iran reverts entirely to covert, proxy efforts to counter US interests, the current US–Iran tensions remain unresolved and will likely continue to persist through the 2020 elections in November.

As a matter of the first order, Soleimani was replaced by his deputy Brigadier General Esmail Ghaani within a day of the former’s death, with Khamenei saying that the Quds Force will be ‘unchanged’.

At the second order, Iraq’s parliament voted in favour of a nonbinding resolution to rescind the invitation to US forces, which led Trump to threaten sanctions and demands for reimbursement. Whether US troops will ultimately leave Iraq (following a ‘mistaken’ report that the US was preparing to depart) remains to be seen, but the destabilization of the US military presence in Iraq fulfils a key Iranian objective.

In the interim, the US-led coalition in Iraq and Syria fighting ISIS announced that it would at least temporarily cease its counterterrorism efforts to instead fortify its outposts and prepare for Iranian retaliation, opening a wider door for the resurgence of the terror group.

By arguing that the US, its troops and interest have not been made safer by Trump’s Middle East policy – from withdrawing from the Iran nuclear deal to the imposition of a ‘maximum pressure campaign’ to Soleimani’s killing – Democrats will be able to point to every post-Soleimani US injury, death, regional terrorism attack, asset compromise, cyberattack and shipping disruption as evidence.

Democratic presidential candidates also ought to be explicit about how they plan to manage tensions with Iran – strategic, diplomatic and military – particularly their position on the future of the nuclear deal.

Iran has made clear that the path to de-escalation is through sanctions relief. Asserting leverage need not always involve taking away all of your counterparty’s options (‘maximum pressure’). It also involves knowing what your adversary wants (sanctions relief) and showing a willingness to offer it (especially where it means less to you) in exchange for something of greater worth (avoiding war/a non-nuclear Iran).

Clarity around future policy of a potential Democratic president may bring de-escalation forward in a way that Trump’s statement of Iran standing down are unlikely to do.




election

Chatham House US 2020 Election Series

A year-long project focused on what is at stake in the pivotal 2020 US presidential and congressional elections, considering the future of US policy on trade, global economy and technology, national security, transatlantic relations, climate, migration and Latin America.

As part of this initiative, we are launching the Chatham House US Foreign Policy Forum as an incubator for foreign policy dialogue and shared expertise outside of the Washington DC framework.

Meeting regularly, in its inaugural year the Forum will largely focus on the 2020 elections, facilitating discussions around developments in the election and critical policy insights.

These activities support the development of a multi-authored volume outlining the current state of play and potential priorities of a second Trump term, as well as a Democratic administration. The text will contribute to the public debate and research considering the resiliency of US institutions as well as the future of US policy engagement abroad.

Situated in London, this project leverages Chatham House’s world-leading, independent foreign policy institute and unparalleled experience in convening multi-stakeholder discussions to provide a unique, international perspective on the 2020 elections.




election

US 2020: Super Tuesday and Implications for the General Election

Invitation Only Research Event

5 March 2020 - 12:00pm to 1:30pm

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

Dr Lindsay Newman, Senior Research Fellow, US and the Americas Programme, Chatham House
Professor Peter Trubowitz, Professor of International Relations, London School of Economics and Political Science; Associate Fellow, US and the Americas Programme, Chatham House
Amy Pope, Associate Fellow, US and the Americas Programme, Chatham House; Deputy Homeland Security Advisor, US National Security Council, 2015-17
Chair: Dr Leslie Vinjamuri, Director, US and the Americas Programme, Chatham House

The US 2020 election season enters a potentially decisive next phase with the Super Tuesday primaries on 3 March. With these fifteen, simultaneously-held state elections, the Democrats hope to have greater clarity about their party’s likely nominee for the general race against President Donald Trump in November. Concerns around intraparty divisions in the Democratic party between progressives (represented by Senators Elizabeth Warren and Bernie Sanders) and moderates (represented by former Vice President Joe Biden and former mayor Pete Buttigieg) have surrounded the primary races so far, and are unlikely to dissipate even if one candidate emerges from the field on 3 March.

Against this backdrop, Chatham House brings together a panel of experts to discuss the state of the Democratic primary race, implications for the general election, and the Trump campaign’s priorities ahead of its re-election bid. Will the Democratic party resolve its divisions and unite behind a progressive or moderate in light of the Super Tuesday election results? How is Trump positioned to fair against the Democratic candidates left in the race? Did Former Mayor of New York Michael Bloomberg’s primary gamble to focus on Super Tuesday pay off? And what policy priorities are likely to be pursued under either a Trump 2.0 or a Democratic administration?

Event attributes

Chatham House Rule

US and Americas Programme




election

Democrats Have Set Themselves Up to Fail in November's Election

21 February 2020

Dr Lindsay Newman

Senior Research Fellow, US and the Americas Programme
Debates and caucuses are proving that the party took the wrong lesson from the midterms. They're now applying that lesson to 2020 with potentially disastrous results.

2020-02-21-DemDebate.jpg

2020 Democratic presidential candidates at the debate in Las Vegas on 19 February. Photo: Getty Images.

The Democratic Party’s struggle for its future policy direction is evident this election season. The primary results in Iowa and New Hampshire, narrow first- and second-place finishes for Senator Bernie Sanders (a progressive) and former South Bend mayor Pete Buttigieg (a moderate), were just two indicators. During Wednesday night’s debate in Las Vegas, the split became even more obvious.

The six candidates onstage clashed on ideology (socialism and capitalism, progressivism and centrism) as well as policy (healthcare, climate change, fossil fuels, criminal justice, China). Buttigieg made plain the stakes for Democrats, saying, 'We’ve got to wake up as a party.'

If a Democratic candidate is elected to be the United States’ 46th president on 3 November, it will be despite this unresolved intra-party struggle.

One lesson the Democratic Party has taken from the 2018 midterm elections is that running candidates across the ideological spectrum is a winning formula.

It is easy to see how they came to this conclusion following the 2016 presidential and 2018 Congressional election experiences. In 2016, the favoured candidate status of former secretary of state Hillary Clinton deterred other aspirants from entering the Democratic primary ahead of a general election she went on to lose to Republican Donald Trump. In 2018, progressive and moderate centrist candidates, both first-timers and incumbents, ran and Democrats retook leadership in the House of Representatives with a 235-seat majority.

But what if this conclusion was noise and not the signal?

The Democratic National Committee (DNC) set the rules for the 2020 election based on the theory that by allowing an inclusive field (more than two dozen candidates entered the presidential race) the campaign processes, including debates, caucuses and primaries, would ultimately identify the most robust, representative candidate to go up against Donald Trump. Perhaps, and somewhat ironically, the 2016 Republican primary process, which involved a wide field culled by Trump’s unexpected success, informed the DNC’s reforms. And while very nice as a hypothesis of Bayesian updating, what has unfolded instead is a scattershot four-way — at times even five-way — race.

In the midst of this party divide, whoever ends up being the Democratic nominee will likely not represent the views of some meaningful proportion of the Democratic base. While healthcare remains the top issue across the Democratic electorate, there are those (candidates and voters) who want a single-payer option for all without a private insurance option and those who want to expand healthcare access while maintaining private insurers. Likewise, on foreign policy, there are those who link US trade policy with protecting American workers and who would therefore continue to use tariffs as a key trade policy, as well as those critical of Trump’s reliance on tariffs.

Compare that with the current state of the Republican Party. Trump’s approval with Republicans is in the high 80s, sometimes even low 90s, and after all but one Republican senator voted to acquit him in the Senate impeachment trial, the party is undeniably Trump’s. A sure sign is the historic turnout for Trump in his essentially uncontested Iowa and New Hampshire primaries.

Their own divisions pose a number of risks, then, for Democrats heading into November’s general election. The first one relates to vulnerabilities arising out of the primary process itself. If the fractures emerging from Iowa and New Hampshire persist, the likelihood of a quick wrap-up of the Democratic primary by April reduces, and the possibility of a contested Democratic convention in July increases (even if from a low base). While exciting television and Twitter fodder, a lengthy primary positions Democrats to go into the fall facing questions of party disunity behind the eventual nominee.

Although complicated to demonstrate empirically, some work has been done to understand whether the protracted 2016 Democratic primary and Sanders’ slow support for Democratic nominee Clinton in 2016 played a part in her defeat and Trump’s electoral success. A delayed general election campaign for the eventual Democratic nominee in 2020 almost certainly advantages President Trump’s money machine, which reportedly has more than twice as much on hand as then-president Barack Obama had going into his 2012 re-election. Further, unlike 2016, which was an open-seat election for the presidency, in 2020 Trump will have a demonstrated incumbent advantage.

The Democratic Party’s succession battle also raises risks around general election turnout. If Sanders is the party’s nominee, Biden or Buttigieg’s constituency may not come out to vote for him. More worrisome for Democrats, if Sanders is the party’s nominee then centrist voters, including those representing the finance industry, may peel off and vote for Trump, who has overseen economic expansion and record unemployment rates following the 2017 tax overhaul and various deregulations.

Alternatively, if Biden, Buttigieg or former mayor Michael Bloomberg become the nominee, Sanders’ many loyal supporters are likely to feel their policy priorities are not represented. And if those voters stay home because the Democratic nominee is not promising a political revolution, evidence suggests that depressed turnout levels may favour Republicans.

A third political peril relates to the business of legislating after the election. If despite the potential pitfalls a Democratic candidate manoeuvres and manages to build a winning coalition on 3 November, they will face the reality of legislative politics, which over the last 10 years have been defined by policy gridlock. Obama managed to get Obamacare through both Democratic-majority congressional chambers, but presided over divided chambers for the remainder of his term. Similarly, Trump’s major legislative accomplishment — the 2017 tax overhaul — was a result of Republican control in both the House of Representatives and the Senate.

A Democratic president will have to make progress on his or her agenda given not only the typical Republican-Democrat divide in Congress, but also facing potential raw divisions within the Democratic Party itself. In such a scenario, a Democratic administration may be tempted to take an expansive view of the president’s authority as we have seen under Trump, including relying on executive actions (tariffs and sanctions) on foreign policy.

The Democratic National Convention in Milwaukee, Wisconsin, beginning 13 July, and the party platform crafted over those four days present an essential opportunity to resolve the party’s divisions before November. If left unchecked, the party might find that its ex ante strategy for the 2020 Democratic primary ends in Trump’s re-election.

This article was originally published in the Independent.




election

COVID-19: America's Looming Election Crisis

8 April 2020

Dr Lindsay Newman

Senior Research Fellow, US and the Americas Programme
Planning now is essential to ensure the legitimacy of November’s elections is not impacted by COVID-19, as vulnerabilities are becoming ever more apparent if voting in person is restricted.

2020-04-08-COVID-US-election

Roadside voting in Madison, Wisconsin in April 2020. Because of coronavirus, the number of polling places was drastically reduced. Photo by Andy Manis/Getty Images.

The COVID-19 epidemic has hit every aspect of American life. The upcoming November general elections will not be immune to the virus’ impact and may be scheduled to happen while the pandemic remains active, or has returned.

There is a danger the epidemic forces change to the way voting takes place this fall, amplifying risks around election security and voter suppression that ultimately undermine the integrity of the elections.

This is further highlighted by the US Supreme Court’s last-minute ruling along ideological lines to restrict an extension on the absentee voting period in the Wisconsin Democratic presidential primary despite the level of infections in the state, forcing voters into a trade-off between their health and their right to vote. The US could be thrown into a political crisis in addition to the health and economic crises it already faces.

Bipartisan sentiment

While France, Chile and Bolivia have already postponed elections in the wake of COVID-19, there is a bipartisan sentiment that the US elections should be held as scheduled on the Tuesday after the first Monday in November. This is enshrined not only in America’s sense of itself – having weathered elections during a civil war, a world war and heightened terrorist alert before – but also in its federal law since 1845.

Despite increasing appetite for federal elections to go ahead in November, there are serious vulnerabilities, which are already becoming visible as connections are drawn between mail-in voting and voter fraud, greater voter access and disadvantages for the Republican party, and city polling closures and Democratic voter suppression.

Concerns around voting access have gained the most attention. If voting in-person is untenable or risky (especially for vulnerable health populations), voters must have alternative means to cast ballots.

During negotiations for the Coronavirus Aid, Relief, and Economic Security (CARES) Act, the Democratic caucus in the House of Representatives proposed $4 billion in state election grants and a nationally-mandated period for early voting and no-excuse absentee voting.

But the final CARES Act sidestepped the access question and stripped funding to $400 million for election security grants to ‘prevent, prepare for, and respond to coronavirus, domestically or internationally, for the 2020 Federal election cycle’. Without knowing exactly what is in store from a cyber-threat perspective, the actual cost for basic election security upgrades is estimated to be $2.1billion. And that is a pre-COVID-19 calculation.

With social-distanced voters likely to be getting more election information than ever from social media, information security is critical to prevent influence from untrustworthy sources. And opportunities for cyber intrusions are likely to increase as states transition to greater virtual registration, plus absentee and mail-in balloting.

This will open new doors on well-documented, existing voter suppression efforts. With the Supreme Court clawing back the Voting Rights Act in 2013 - allowing certain states to make changes to election and voting laws without federal pre-clearance - heightened election security requirements, such as exact match campaigns and voter purges, have been used to justify voter suppression.

As more vote remotely in the remaining primaries (many now rescheduled for 2 June) and the November general elections, the added burden on states around verification will only increase temptation to set aside ‘non-compliant’ ballots. Especially as some in the Republican Party, including Donald Trump, have advocated a contested view that higher turnout favours the Democratic Party.

A fundamental principle of US democracy is that losers of elections respect the result, but history shows that election results have been contested. In 2000, it took weeks for a result to be confirmed in the presidential election. More recently, in the 2018 race for governor in Georgia, allegations of voter suppression raised questions about the validity of the eventual result.

Without proper access, security, and verification the electoral process – whenever it takes place – will become vulnerable to questions of integrity. The federal response to the initial spread of COVID-19 saw costly delays which pushed the US into a public health crisis and economic contraction.

Any narrative thread of election illegitimacy with November’s elections will further pull apart the fabric of a country already frayed by coronavirus. Federal and state authorities must start planning now for how the US will hold elections in the midst - or immediate aftermath - of COVID-19.




election

Three Takeaways From the Belarusian Parliamentary Elections

28 November 2019

Ryhor Astapenia

Robert Bosch Stiftung Academy Fellow, Russia and Eurasia Programme
Lukashenka’s domestic support is waning and he is not willing to make concessions to the West. Instead, he is trying to appease the ruling cadre.

2019-11-28-Luk.jpg

Alexander Lukashenka leaves a voting booth on 17 November. Photo: Getty Images.

Belarus’s parliamentary elections, held on 17 November, were predictably non-transparent, with numerous violations. The regime of Alexander Lukasheka allowed no opposition candidates as members of parliament – in contrast to the previous parliament, in which there were two opposition MPs. While this might seem to be a return to ‘business as usual’, three key takeaways from the elections highlight a shifting political and social landscape.

1. Lukashenka is appeasing his ruling cadre by promising to increase their role in the political system.

With several influential officials becoming new MPs, it is more likely that parliament will be more involved in any forthcoming discussion of a new constitution. Lukashenka has been promising constitutional reform for several years; he has said publicly that it will lead to an increased significance of government agencies as well as parliament. The aim of this is to keep them more engaged and on Lukashenka’s side.

In terms of the composition of the new parliament itself, there are some key differences with previous years. It is no longer a comfortable place for officials to while away their pre-retirement: many MPs are now in their fifties or younger, and have plans for careers beyond parliament.

It also looks as if small steps are being taken towards the emergence of a party system in Belarus. The leader of Belaya Rus, a pro-government association of Belarusian officials, got a seat in parliament for the first time, increasing the likelihood of it becoming a political party. The number of MPs from different parties has increased to 21 (out of 110 in total). Although these still all broadly support Lukashenka, they can differ from the president in policy positions. For example, the Labour and Justice Party, with 6 seats in parliament, supported the annexation of Crimea in 2014. Change, of course, may not necessarily be in a pro-Western direction.

Moreover, the newly elected members of parliament look more like real politicians. They go to debates, speak to independent journalists and have their own social media channels. Some have even felt able to criticize the actions of the authorities.

Similar changes have been taking place in other institutions in Belarus. The government is now more competent than it has ever been. The National Bank has managed to carry out macroeconomic stabilization on one of the most unstable currencies in Europe, while the ministries responsible for the economic development have implemented certain small-scale reforms. The Minister of the Interior has even acknowledged mistakes made by his department (under his predecessor), and undertaken to make improvements.

This has resulted in a near-comical situation, whereby the Belarusian non-state media outlets have an increasingly positive view of some state officials, such as Prime Minister Siarhei Rumas, while the state media has been scaling back its coverage of him to ensure he does not become too popular.

2. Belarus has less need for the West and is reluctant to make even small concessions.

Since the slight warming of Belarusian relations with the West in 2014, Lukashenka has been having more meetings with prominent Western officials. Western institutions began trying to cooperate more closely with Belarus, but soon saw that it was not very interested. In 2018, the European Bank of Reconstruction and Development allocated €360 million to Belarus. However, it is now considering a reduction, as reforms in Belarus have not, in its view, gone far enough. The European Union has also committed considerable time and money to regional authorities, but this has not brought any significant changes to Belarusian local government.

The absence of opposition candidates also demonstrates that the Belarusian authorities are prepared for a new deterioration of relations with the West. The authorities could certainly have afforded some opposition in parliament, especially since they themselves choose whom to appoint. Given that they did not, they either do not consider worsened relations a problem or are confident that the West will continue to cooperate with Belarus in order to limit Russian power, regardless of its violations of the rule of law.

3. There is growing popular dissatisfaction with the current regime, but the state has no good plan for how to deal with it.

Parliamentary election campaigns in Belarus are traditionally low-key, but this year they were particularly muted. The authorities tried to ensure that people knew as little about the election as possible. Campaign posters appeared on the streets just two weeks before polling day. It seems the authorities were reluctant to politicize society, as further resentment at autocratic rule is brewing.

Many Belarusians who previously supported Lukashenka now have a very critical opinion of him. Take political blogging: the most popular political blogger in Belarus is a 22-year-old man who goes by the name of NEXTA. He produces low-quality videos which are highly critical of the authorities. A film by him about Lukashenka, released a month ago, has already received 1.8 million views, even though there are only 9.5 million people in Belarus.

The authorities are not in a concessionary mood. The presidential elections in 2020 will also likely be a sham. If the authorities’ grip over the country is weakened, they will fear an outbreak of anger, resulting in widespread protests which the regime might once again have to meet with violence.




election

Trudeau Election Marks New Start for Canada

22 October 2015

Cleo Paskal

Associate Fellow, Energy, Environment and Resources Programme and Asia-Pacific Programme
Canada's partners are about to see a sea change in policy from the new Liberal government on a range of issues, especially relating to energy and the environment.

20151022Trudeau.jpg

Canadian Liberal Party leader Justin Trudeau arrives to give a press conference in Ottawa on 20 October 2015. Photo by Getty Images.

The Liberal Party likes to think of itself as ‘Canada’s natural governing party’. Though they have been out of power for 10 years, and newly elected prime minister Justin Trudeau is relatively inexperienced, the party has deep and old networks across the country. The backroom is full of experienced old hands, including former prime ministers who have known Trudeau since he was a toddler. With a number of bold shifts promised in Canadian foreign policy, and a willingness to be fiscally expansive (Trudeau has said he is willing to run deficits for three years to implement their policies), the change is likely to be dramatic, fast and unrelenting.

International engagement

The Liberals’ proposed new Canada has a hint of nostalgia for a time when Canadians thought of themselves as a force for good in the world. There is likely to be a large-scale reengagement with the international community, especially the United Nations, not only on climate policy, but also peacekeeping and disaster response. Canada’s French-speaking experts make it well placed to engage in crisis-hit French-speaking countries while carrying somewhat less colonial baggage. Meanwhile, the Liberals have promised to withdraw Canadian forces from the combat mission against ISIS in Iraq, pledging instead to reorient their focus towards aid.

Science-based policies

Trudeau has the backing of a very deep bench of experienced parliamentarians, some of whom have been waiting a long time to get back in the governing game. They include Stéphane Dion, the former environment minister who gained respect for his chairing of the 2005 UN climate change conference in Montreal, and Marc Garneau, former astronaut and president of the Canadian Space Agency. This will be a government that understands and values (and has promised to restore funding for) scientific research.

This means more investment in climate resilience, renewables and other forward planning policies, but also a reexamination of some of Canada’s long-standing strengths, including Canada’s fisheries. Under the previous government, over half-a-dozen critical fisheries research libraries were shut down, in some cases with large-scale loss of data. A more science-based fisheries policy, combined with the promised funding to the Royal Canadian Navy, may result in a reinvigorated and coordinated Canadian fisheries policy. This would become particularly important in a time of global food constraints.

‘Nation-to-nation’

Trudeau’s stated goal of establishing ‘nation-to-nation’ relations with Canada’s indigenous peoples has potentially global implications.

Through treaties, Canada’s First Nations can stake a claim to about a third of Canada’s landmass, including resource rich areas. They own or control access to land that contains oil, gas, uranium, gold, diamonds and much more. China, for one, realized the potential power of First Nations as far back as 2008, when Beijing invited over two dozen indigenous leaders to China to talk business. During that trip, Peguis Chief Glenn Hudson explained that the trip was ‘an important step for us in moving forward. Our future is not only in Canada, but partnering with other countries.’

Trudeau’s ‘nation-to-nation’ statement acknowledges the strengthening role of First Nations in Canada’s resource policy. Indigenous communities have been among the most adversely affected by oil sands development and, in other parts of Canada, have successfully blocked resources extraction or transit. For example, in May, the Lax Kw'alaams Band refused to allow a proposed multi-million dollar Petronas LNG project to be developed on their land over concerns it would affect their economically and culturally important salmon runs.

Real First Nations engagement with resource management has the potential substantially change Canada’s energy, environment and resources profile. First Nations in eastern Canada have already blocked hydro development that could power much of the northeastern United States and, in a time of pipeline expansion, First Nations might become instrumental in deciding if more pipelines will run north-to-south, towards the border with the United States, or east-to-west, towards the Pacific and the markets in Asia.

Energy sector

At the same time, while Trudeau said he was against the Northern Gateway pipeline, in large part because of how it would affect the people of the Pacific coast, he backed Keystone XL. However, he is unlikely to push for it in the face of a veto by President Obama. Also, low fossil fuel prices contributed to the election in May of a left-leaning provincial government in the badly hit oil heartland of Alberta. If prices stay low, that could very well combine with Liberal campaign promises to put a coordinated national price on carbon and to phase out fossil fuel subsidies, additionally reshaping Canada’s energy landscape..

Agriculture and the Trans-Pacific Partnership

The Liberals also have a stated goal of investing in a more innovative and safe agricultural sector, while defending the interests of Canadian farmers. This dovetails with their promise to openly examine and discuss the Trans-Pacific Partnership (TPP) before ratification. Canadian dairy farmers in particular were concerned about some of the TPP provisions but, without access to the details of the text, it was difficult to estimate how the sector would be affected.

One proposed agricultural policy that may have widespread economic implications is the legalization, regulation and taxation of marijuana. As some states in the United States have found, this has the potential to be economically beneficial and, for Canada, a possible major boon to cross-border tourism.

In a myriad of small ways, and a few big ones, the new Liberal government plans to change Canada’s international role and domestic landscape. This is one to watch.

To comment on this article, please contact Chatham House Feedback 




election

Behind the campaign promises - what the NHS means for the election

UK general election has been called - polling day is on the 12th of December, and from now until then we’re going to be bringing you a weekly election-themed podcast. We want to help you make sense of the promises and pledges, claims and counter-claims, that are being made around healthcare and the NHS out on the campaign trail. This week we're...




election

Fuel selection in human skeletal muscle in insulin resistance: a reexamination

DE Kelley
May 1, 2000; 49:677-683
Articles




election

Webinar: The UK's Unpredictable General Election?

Members Event Webinar

19 November 2019 - 11:30am to 12:00pm

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

Professor Matthew Goodwin, Visiting Senior Fellow, Europe Programme, Chatham House
Professor David Cutts, Associate Fellow, Europe Programme, Chatham House

On 12 December 2019, the United Kingdom will go to the polls in a fifth nationwide vote in only four years. This is expected to be one of the most unpredictable general elections in the nation’s post-war history with the anti-Brexit Liberal Democrats and Nigel Farage’s Eurosceptic Brexit Party both presenting a serious challenge to the UK’s established two-party system.

This webinar will discuss the UK general election and will unpack some of the reasons behind its supposed unpredictability. To what extent will this be a Brexit election and what are the other issues at the forefront of voters’ minds? And will the outcome of the election give us a clear indication of the UK’s domestic, European and wider international political trajectory?

Please note, this event is online only. Members can watch webinars from a computer or another internet-ready device and do not need to come to Chatham House to attend.




election

Reflections on the Brexit Election

Invitation Only Research Event

6 December 2019 - 8:30am to 9:30am

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

Alistair Burt, Conservative Member of Parliament (1983-97 and 2001-19); Minister of State for the Middle East, UK Foreign & Commonwealth Office and Minister of State at the Department for UK International Development (2017-19)

On 12 December 2019, the United Kingdom will hold one of its most crucial elections in the 21st century. The result will have a direct impact on the Brexit process and will most likely determine the country’s future direction for years to come.  

Yet the final outcome is far from predictable. It seems quite certain that the 2019 election is unlikely to produce a clear two-party share of the vote as happened back in 2017. Public trust in politicians is low and party loyalty is looser than ever. Polls show that Brexit it overwhelmingly considered as the main issue among the electorate alongside a deep concern about the future of public services. This raises multiple questions: can the 2019 election represent a chance to unite the country and move on? Will cross-party identities of ‘Leavers’ and ‘Remainers’ translate to how people vote in the election? And what will the outcome mean for Brexit and the future of party politics in Britain? 

Attendance at this event is by invitation only. 

Event attributes

Chatham House Rule

Alina Lyadova

Europe Programme Coordinator




election

UK General Election 2019: BBC-Chatham House Foreign Policy Debate

Members Event

28 November 2019 - 10:30am to 11:30am

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

Stephen Gethins, Candidate for Fife North East and Shadow Spokesperson for Foreign and Commonwealth Office (2018-19), Scottish National Party (remote)
Dominic Raab, Candidate for Esher & Walton, Foreign Secretary and First Secretary of State (2019), Conservative Party
Emily Thornberry, Candidate for Islington South & Finsbury and Shadow Foreign Secretary (2016-19), Labour Party
Chuka Umunna, Candidate for Cities of London & Westminster and Spokesperson for Foreign and Commonwealth Affairs (2019), Liberal Democrat Party
Chair: Ritula Shah, The World Tonight, BBC Radio 4
 

As the United Kingdom prepares to go to the polls on 12 December 2019, this event, organized in conjunction with the BBC's show, The World Tonight, will give a Chatham House audience the opportunity to put their foreign policy questions to a panel of senior UK politicians.
 
Important Information About the Event
This event is hosted in collaboration with the BBC. It will be recorded and broadcast on BBC Radio 4The World Tonight. Given the BBC’s commitment to due impartiality during the election period, questions will be asked to be pre-submitted via email by audience members the day before the event. The BBC will then select a balanced and diverse range of questions from those submitted. The panellists will not receive the questions in advance of the event.
 
About the Ballot
Due to the expected popularity of this event, this event will be balloted. Please register your interest for the event by 11:59pm on Monday 25 November. Successful registrants – selected at random – will be notified on Tuesday 26 November and then invited to submit their questions. The BBC will contact directly those audience members whose questions they select.
 
As priority will be given to members, we are unable to register members’ guests for this event.

Event attributes

E-ticket event

Members Events Team




election

UK General Election 2019: Foreign Policy Implications

Members Event

11 December 2019 - 5:30pm to 6:30pm

Chatham House | 10 St James's Square | London | SW1Y 4LE

Event participants

John Casson, Associate Fellow, Europe Programme, Chatham House
Tom Raines, Head, Europe Programme, Chatham House
Dr Yu Jie, Senior Research Fellow, Asia-Pacific Programme, Chatham House
Antony Froggatt, Senior Research Fellow, Energy, Environment and Resource Governance, Chatham House

On the eve of the UK general election, a group of senior Chatham House experts analyse a range of key foreign policy issues that will hinge on the election’s outcome.

Members Events Team




election

Immigration under Trump: A Review of Policy Shifts in the Year Since the Election

On the campaign trail, Donald Trump made immigration the centerpiece of his campaign, offering a more detailed policy agenda than on any other issue. In the year since the election that propelled the Republican into the White House, how has the Trump administration’s record matched up with the rhetoric? This policy brief examines the executive orders and other changes to existing policy and practice made during 2017.




election

"Merit-Based" Immigration: Designing Successful Selection Systems

With the U.S. administration calling for the United States to adopt a more “merit-based” immigrant selection system, this conversation focused on what policymakers should consider in designing—and managing—immigrant selection systems in a time of intense labor-market and demographic change.




election

Poland postpones its presidential election

Poland announced a delay of its Sunday presidential election, Europe's first planned vote during the coronavirus pandemic.




election

"Merit-Based" Immigration: Designing Successful Selection Systems

MPI and OECD experts discuss what policymakers should consider in designing and managing immigrant selection systems in a time of intense labor-market and demographic change.




election

“Merit-Based” Immigration: Trump Proposal Would Dramatically Revamp Immigrant Selection Criteria, But with Modest Effects on Numbers

The Trump administration’s plan to create a "merit-based" U.S. immigration system, lessening the longstanding focus on family reunification in favor of more economic migrants, has met with a lackluster response from Democrats and Republicans alike. This Policy Beat article explores how the Trump proposal would reshape immigration to the United States, and how it compares to selection systems in other countries and past debates about changing the U.S. system.




election

The Dutch Elections: How to Lose and Still Shape the Direction of a Country—and Possibly a Continent?

The failure of Geert Wilders’ right-wing, anti-Islam Freedom Party (PVV) to become the top vote-getter in the Dutch parliamentary elections is being hailed as proof of the limits of anti-Muslim rhetoric and even the “waning” of the appeal of right-wing populism. But as this commentary explores, a closer reading leads one to a more nuanced interpretation of the results and the recognition that Wilders will remain a major force.




election

Equipping Immigrant Selection Systems for a Changing World of Work (Transatlantic Council Statement)

As technological developments—from automation to artificial intelligence and machine learning—reshape the world of work, governments face the challenge of updating how they attract, select, and retain economic-stream immigrants. This report, concluding a series on building migration systems for a new age of economic competitiveness, lays out the key considerations for "future-proofing" immigrant selection systems.




election

Swing State: Immigration's Impact on the 2020 U.S. Election

This event will feature a screening of the short documentary The Fields of Immokalee, and include a discussion on immigration and politics in Florida and other swing states in the run-up to the 2020 election. 




election

The Decoy Effect, or How to Win an Election

If Democrats Hillary Clinton and Barack Obama ever took a break from fundraising to bone up on psychology, they might realize the need to talk up . . . John Edwards.




election

Rules About Delegates Can Sway an Election

Sen. John McCain's quest for the Republican presidential nomination was once seen as dead, but like those robots in the "Terminator" movies that reassemble themselves after being blown to smithereens, he came back. Five years ago, Sen. Barack Obama (Ill.) was a virtually unknown African American ...




election

Utah Ruling Highlights Sticky Issue of Partisan School Board Elections

In Florida, Indiana, Kansas, North Carolina, Tennessee and Utah, Republicans are arguing that candidates for local and state school boards should run on party tickets.




election

Election Night Brings Highs and Lows for Oklahoma Teachers

About a dozen teachers running for state legislative seats gathered with their supporters in Tulsa for a watch party on election night. The results were better for some than others.




election

2018 Election Will Rock California Education

Next year's California election will be both a referendum on the massive changes in education finance and testing enacted during Jerry Brown governorship and a test of the political coalition that made those changes possible.




election

Charter Advocates Dealt Loss in California Chief's Election

State Assembly member Tony Thurmond ultimately prevailed over former charter school executive Marshall Tuck in a contest that drew more than $50 million in outside campaign spending.




election

What Are the K-12 Policy Stakes in N.J. and Virginia Elections?

Education policy analysts are closely watching Tuesday's races for governor and state legislature in both states to see what messages about K-12 could resonate when many more states hold elections next year.




election

Utah Ruling Highlights Sticky Issue of Partisan School Board Elections

In Florida, Indiana, Kansas, North Carolina, Tennessee and Utah, Republicans are arguing that candidates for local and state school boards should run on party tickets.




election

Education Advocates Already Filing to Run in 2018 State Elections

Already, some educators and prominent education advocates have entered their names into the running for of the many 2018 state races around the country where education policy is likely to be a hot topic.




election

Periodicals archive online : JISC collections selection 2

This is a subset of full text journal backfiles in the arts, humanities, and social sciences from the Periodicals Archive Online collection, dating from 1891-.




election

The descent of man, and selection in relation to sex / by Charles Darwin.

London : J. Murray, 1894.




election

Directions for drinking the Cheltenham waters : with a selection of cases, illustrating their effects in a great variety of diseases / by James M'Cabe.

Cheltenham : printed for G.A. Williams, librarian, [1823]




election

To Show That Elections Matter, This Teacher Is Running for Office

In a civics lesson come to life, this Missouri high school government teacher is running for state legislature.




election

Consistent model selection criteria and goodness-of-fit test for common time series models

Jean-Marc Bardet, Kare Kamila, William Kengne.

Source: Electronic Journal of Statistics, Volume 14, Number 1, 2009--2052.

Abstract:
This paper studies the model selection problem in a large class of causal time series models, which includes both the ARMA or AR($infty $) processes, as well as the GARCH or ARCH($infty $), APARCH, ARMA-GARCH and many others processes. To tackle this issue, we consider a penalized contrast based on the quasi-likelihood of the model. We provide sufficient conditions for the penalty term to ensure the consistency of the proposed procedure as well as the consistency and the asymptotic normality of the quasi-maximum likelihood estimator of the chosen model. We also propose a tool for diagnosing the goodness-of-fit of the chosen model based on a Portmanteau test. Monte-Carlo experiments and numerical applications on illustrative examples are performed to highlight the obtained asymptotic results. Moreover, using a data-driven choice of the penalty, they show the practical efficiency of this new model selection procedure and Portemanteau test.




election

A fast and consistent variable selection method for high-dimensional multivariate linear regression with a large number of explanatory variables

Ryoya Oda, Hirokazu Yanagihara.

Source: Electronic Journal of Statistics, Volume 14, Number 1, 1386--1412.

Abstract:
We put forward a variable selection method for selecting explanatory variables in a normality-assumed multivariate linear regression. It is cumbersome to calculate variable selection criteria for all subsets of explanatory variables when the number of explanatory variables is large. Therefore, we propose a fast and consistent variable selection method based on a generalized $C_{p}$ criterion. The consistency of the method is provided by a high-dimensional asymptotic framework such that the sample size and the sum of the dimensions of response vectors and explanatory vectors divided by the sample size tend to infinity and some positive constant which are less than one, respectively. Through numerical simulations, it is shown that the proposed method has a high probability of selecting the true subset of explanatory variables and is fast under a moderate sample size even when the number of dimensions is large.




election

On the distribution, model selection properties and uniqueness of the Lasso estimator in low and high dimensions

Karl Ewald, Ulrike Schneider.

Source: Electronic Journal of Statistics, Volume 14, Number 1, 944--969.

Abstract:
We derive expressions for the finite-sample distribution of the Lasso estimator in the context of a linear regression model in low as well as in high dimensions by exploiting the structure of the optimization problem defining the estimator. In low dimensions, we assume full rank of the regressor matrix and present expressions for the cumulative distribution function as well as the densities of the absolutely continuous parts of the estimator. Our results are presented for the case of normally distributed errors, but do not hinge on this assumption and can easily be generalized. Additionally, we establish an explicit formula for the correspondence between the Lasso and the least-squares estimator. We derive analogous results for the distribution in less explicit form in high dimensions where we make no assumptions on the regressor matrix at all. In this setting, we also investigate the model selection properties of the Lasso and show that possibly only a subset of models might be selected by the estimator, completely independently of the observed response vector. Finally, we present a condition for uniqueness of the estimator that is necessary as well as sufficient.




election

Modal clustering asymptotics with applications to bandwidth selection

Alessandro Casa, José E. Chacón, Giovanna Menardi.

Source: Electronic Journal of Statistics, Volume 14, Number 1, 835--856.

Abstract:
Density-based clustering relies on the idea of linking groups to some specific features of the probability distribution underlying the data. The reference to a true, yet unknown, population structure allows framing the clustering problem in a standard inferential setting, where the concept of ideal population clustering is defined as the partition induced by the true density function. The nonparametric formulation of this approach, known as modal clustering, draws a correspondence between the groups and the domains of attraction of the density modes. Operationally, a nonparametric density estimate is required and a proper selection of the amount of smoothing, governing the shape of the density and hence possibly the modal structure, is crucial to identify the final partition. In this work, we address the issue of density estimation for modal clustering from an asymptotic perspective. A natural and easy to interpret metric to measure the distance between density-based partitions is discussed, its asymptotic approximation explored, and employed to study the problem of bandwidth selection for nonparametric modal clustering.




election

DESlib: A Dynamic ensemble selection library in Python

DESlib is an open-source python library providing the implementation of several dynamic selection techniques. The library is divided into three modules: (i) dcs, containing the implementation of dynamic classifier selection methods (DCS); (ii) des, containing the implementation of dynamic ensemble selection methods (DES); (iii) static, with the implementation of static ensemble techniques. The library is fully documented (documentation available online on Read the Docs), has a high test coverage (codecov.io) and is part of the scikit-learn-contrib supported projects. Documentation, code and examples can be found on its GitHub page: https://github.com/scikit-learn-contrib/DESlib.




election

Robust Bayesian model selection for heavy-tailed linear regression using finite mixtures

Flávio B. Gonçalves, Marcos O. Prates, Victor Hugo Lachos.

Source: Brazilian Journal of Probability and Statistics, Volume 34, Number 1, 51--70.

Abstract:
In this paper, we present a novel methodology to perform Bayesian model selection in linear models with heavy-tailed distributions. We consider a finite mixture of distributions to model a latent variable where each component of the mixture corresponds to one possible model within the symmetrical class of normal independent distributions. Naturally, the Gaussian model is one of the possibilities. This allows for a simultaneous analysis based on the posterior probability of each model. Inference is performed via Markov chain Monte Carlo—a Gibbs sampler with Metropolis–Hastings steps for a class of parameters. Simulated examples highlight the advantages of this approach compared to a segregated analysis based on arbitrarily chosen model selection criteria. Examples with real data are presented and an extension to censored linear regression is introduced and discussed.




election

Variable selection methods for model-based clustering

Michael Fop, Thomas Brendan Murphy.

Source: Statistics Surveys, Volume 12, 18--65.

Abstract:
Model-based clustering is a popular approach for clustering multivariate data which has seen applications in numerous fields. Nowadays, high-dimensional data are more and more common and the model-based clustering approach has adapted to deal with the increasing dimensionality. In particular, the development of variable selection techniques has received a lot of attention and research effort in recent years. Even for small size problems, variable selection has been advocated to facilitate the interpretation of the clustering results. This review provides a summary of the methods developed for variable selection in model-based clustering. Existing R packages implementing the different methods are indicated and illustrated in application to two data analysis examples.




election

Errata: A survey of Bayesian predictive methods for model assessment, selection and comparison

Aki Vehtari, Janne Ojanen.

Source: Statistics Surveys, Volume 8, , 1--1.

Abstract:
Errata for “A survey of Bayesian predictive methods for model assessment, selection and comparison” by A. Vehtari and J. Ojanen, Statistics Surveys , 6 (2012), 142–228. doi:10.1214/12-SS102.




election

A survey of Bayesian predictive methods for model assessment, selection and comparison

Aki Vehtari, Janne Ojanen

Source: Statist. Surv., Volume 6, 142--228.

Abstract:
To date, several methods exist in the statistical literature for model assessment, which purport themselves specifically as Bayesian predictive methods. The decision theoretic assumptions on which these methods are based are not always clearly stated in the original articles, however. The aim of this survey is to provide a unified review of Bayesian predictive model assessment and selection methods, and of methods closely related to them. We review the various assumptions that are made in this context and discuss the connections between different approaches, with an emphasis on how each method approximates the expected utility of using a Bayesian model for the purpose of predicting future data.




election

A survey of cross-validation procedures for model selection

Sylvain Arlot, Alain Celisse

Source: Statist. Surv., Volume 4, 40--79.

Abstract:
Used to estimate the risk of an estimator or to perform model selection, cross-validation is a widespread strategy because of its simplicity and its (apparent) universality. Many results exist on model selection performances of cross-validation procedures. This survey intends to relate these results to the most recent advances of model selection theory, with a particular emphasis on distinguishing empirical statements from rigorous theoretical results. As a conclusion, guidelines are provided for choosing the best cross-validation procedure according to the particular features of the problem in hand.




election

Additive Bayesian variable selection under censoring and misspecification. (arXiv:1907.13563v3 [stat.ME] UPDATED)

We study the interplay of two important issues on Bayesian model selection (BMS): censoring and model misspecification. We consider additive accelerated failure time (AAFT), Cox proportional hazards and probit models, and a more general concave log-likelihood structure. A fundamental question is what solution can one hope BMS to provide, when (inevitably) models are misspecified. We show that asymptotically BMS keeps any covariate with predictive power for either the outcome or censoring times, and discards other covariates. Misspecification refers to assuming the wrong model or functional effect on the response, including using a finite basis for a truly non-parametric effect, or omitting truly relevant covariates. We argue for using simple models that are computationally practical yet attain good power to detect potentially complex effects, despite misspecification. Misspecification and censoring both have an asymptotically negligible effect on (suitably-defined) false positives, but their impact on power is exponential. We portray these issues via simple descriptions of early/late censoring and the drop in predictive accuracy due to misspecification. From a methods point of view, we consider local priors and a novel structure that combines local and non-local priors to enforce sparsity. We develop algorithms to capitalize on the AAFT tractability, approximations to AAFT and probit likelihoods giving significant computational gains, a simple augmented Gibbs sampler to hierarchically explore linear and non-linear effects, and an implementation in the R package mombf. We illustrate the proposed methods and others based on likelihood penalties via extensive simulations under misspecification and censoring. We present two applications concerning the effect of gene expression on colon and breast cancer.




election

Feature Selection Methods for Uplift Modeling. (arXiv:2005.03447v1 [cs.LG])

Uplift modeling is a predictive modeling technique that estimates the user-level incremental effect of a treatment using machine learning models. It is often used for targeting promotions and advertisements, as well as for the personalization of product offerings. In these applications, there are often hundreds of features available to build such models. Keeping all the features in a model can be costly and inefficient. Feature selection is an essential step in the modeling process for multiple reasons: improving the estimation accuracy by eliminating irrelevant features, accelerating model training and prediction speed, reducing the monitoring and maintenance workload for feature data pipeline, and providing better model interpretation and diagnostics capability. However, feature selection methods for uplift modeling have been rarely discussed in the literature. Although there are various feature selection methods for standard machine learning models, we will demonstrate that those methods are sub-optimal for solving the feature selection problem for uplift modeling. To address this problem, we introduce a set of feature selection methods designed specifically for uplift modeling, including both filter methods and embedded methods. To evaluate the effectiveness of the proposed feature selection methods, we use different uplift models and measure the accuracy of each model with a different number of selected features. We use both synthetic and real data to conduct these experiments. We also implemented the proposed filter methods in an open source Python package (CausalML).




election

lmSubsets: Exact Variable-Subset Selection in Linear Regression for R

An R package for computing the all-subsets regression problem is presented. The proposed algorithms are based on computational strategies recently developed. A novel algorithm for the best-subset regression problem selects subset models based on a predetermined criterion. The package user can choose from exact and from approximation algorithms. The core of the package is written in C++ and provides an efficient implementation of all the underlying numerical computations. A case study and benchmark results illustrate the usage and the computational efficiency of the package.




election

Uniformly valid confidence intervals post-model-selection

François Bachoc, David Preinerstorfer, Lukas Steinberger.

Source: The Annals of Statistics, Volume 48, Number 1, 440--463.

Abstract:
We suggest general methods to construct asymptotically uniformly valid confidence intervals post-model-selection. The constructions are based on principles recently proposed by Berk et al. ( Ann. Statist. 41 (2013) 802–837). In particular, the candidate models used can be misspecified, the target of inference is model-specific, and coverage is guaranteed for any data-driven model selection procedure. After developing a general theory, we apply our methods to practically important situations where the candidate set of models, from which a working model is selected, consists of fixed design homoskedastic or heteroskedastic linear models, or of binary regression models with general link functions. In an extensive simulation study, we find that the proposed confidence intervals perform remarkably well, even when compared to existing methods that are tailored only for specific model selection procedures.




election

Consistent selection of the number of change-points via sample-splitting

Changliang Zou, Guanghui Wang, Runze Li.

Source: The Annals of Statistics, Volume 48, Number 1, 413--439.

Abstract:
In multiple change-point analysis, one of the major challenges is to estimate the number of change-points. Most existing approaches attempt to minimize a Schwarz information criterion which balances a term quantifying model fit with a penalization term accounting for model complexity that increases with the number of change-points and limits overfitting. However, different penalization terms are required to adapt to different contexts of multiple change-point problems and the optimal penalization magnitude usually varies from the model and error distribution. We propose a data-driven selection criterion that is applicable to most kinds of popular change-point detection methods, including binary segmentation and optimal partitioning algorithms. The key idea is to select the number of change-points that minimizes the squared prediction error, which measures the fit of a specified model for a new sample. We develop a cross-validation estimation scheme based on an order-preserved sample-splitting strategy, and establish its asymptotic selection consistency under some mild conditions. Effectiveness of the proposed selection criterion is demonstrated on a variety of numerical experiments and real-data examples.