To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure email@example.com
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Site-specific weed management (on the scale of a few meters or less) has the potential to greatly reduce pesticide use and its associated environmental and economic costs. A prerequisite for site-specific weed management is the availability of accurate maps of the weed population that can be generated quickly and cheaply. Improvements and cost reductions in unmanned aerial vehicles (UAVs) and camera technology mean these tools are now readily available for agricultural use. We used UAVs to collect aerial images captured in both RGB and multispectral formats of 12 cereal fields (wheat [Triticum aestivum L.] and barley [Hordeum vulgare L.]) across eastern England. These data were used to train machine learning models to generate prediction maps of locations of black-grass (Alopecurus myosuroides Huds.), a prolific weed in UK cereal fields. We tested machine learning and data set resampling methods to obtain the most accurate system for predicting the presence and absence of weeds in new out-of-sample fields. The accuracy of the system in predicting the absence of A. myosuroides is 69% and its presence above 5 g in weight with 77% accuracy in new out-of-sample fields. This system generates prediction maps that can be used by either agricultural machinery or autonomous robotic platforms for precision weed management. Improvements to the accuracy can be made by increasing the number of fields and samples in the data set and the length of time over which data are collected to gather data across the entire growing season.
To explore patterns of post-malnutrition growth (PMGr) during and after treatment for severe malnutrition and describe associations with survival and non-communicable disease (NCD) risk 7 years post-treatment.
Six indicators of PMGr were derived based on a variety of timepoints, weight, weight-for-age z-score and height-for-age z-score (HAZ). Three categorisation methods included no categorisation, quintiles and latent class analysis (LCA). Associations with mortality risk and seven NCD indicators were analysed.
Secondary data from Blantyre, Malawi between 2006 and 2014.
A cohort of 1024 children treated for severe malnutrition (weight-for-length z-score < 70 % median and/or MUAC (mid-upper arm circumference) < 110 mm and/or bilateral oedema) at ages 5–168 months.
Faster weight gain during treatment (g/d) and after treatment (g/kg/day) was associated with lower risk of death (adjusted OR 0·99, 95 % CI 0·99, 1·00; and adjusted OR 0·91, 95 % CI 0·87, 0·94, respectively). In survivors (mean age 9 years), it was associated with greater hand grip strength (0·02, 95 % CI 0·00, 0·03) and larger HAZ (6·62, 95 % CI 1·31, 11·9), both indicators of better health. However, faster weight gain was also associated with increased waist:hip ratio (0·02, 95 % CI 0·01, 0·03), an indicator of later-life NCD risk. The clearest patterns of association were seen when defining PMGr based on weight gain in g/d during treatment and using the LCA method to describe growth patterns. Weight deficit at admission was a major confounder.
A complex pattern of benefits and risks is associated with faster PMGr. Both initial weight deficit and rate of weight gain have important implications for future health.
The domestic dog is the reservoir host of Leishmania infantum, the causative agent of zoonotic visceral leishmaniasis endemic in Mediterranean Europe. Targeted control requires predictive risk maps of canine leishmaniasis (CanL), which are now explored. We databased 2187 published and unpublished surveys of CanL in southern Europe. A total of 947 western surveys met inclusion criteria for analysis, including serological identification of infection (504, 369 dogs tested 1971–2006). Seroprevalence was 23 2% overall (median 10%). Logistic regression models within a GIS framework identified the main environmental predictors of CanL seroprevalence in Portugal, Spain, France and Italy, or in France alone. A 10-fold cross-validation approach determined model capacity to predict point-values of seroprevalence and the correct seroprevalence class (<5%, 5–20%, >20%). Both the four-country and France-only models performed reasonably well for predicting correctly the <5% and >20% seroprevalence classes (AUC >0 70). However, the France-only model performed much better for France than the four-country model. The four-country model adequately predicted regions of CanL emergence in northern Italy (<5% seroprevalence). Both models poorly predicted intermediate point seroprevalences (5–20%) within regional foci, because surveys were biased towards known rural foci and Mediterranean bioclimates. Our recommendations for standardizing surveys would permit higher-resolution risk mapping.
Cox and Katz argue that the massive wave of redistricting that occurred post-1964 created an “incumbency advantage” in the U.S. House. They find that the political composition of the courts and state legislatures that redrew the districts are critical to understanding which party benefited from redistricting. On balance, they find that the redistricting created a larger advantage for Republican incumbents.
THE REAPPORTIONMENT REVOLUTION
A Sketch of the Reapportionment Revolution
The Court's Decisions
On March 26, 1962, the Supreme Court handed down its decision in the case of Baker v. Carr, thus initiating what has since been known as the reapportionment revolution. The suit was brought by urban plaintiffs in Tennessee, who challenged their state legislature's failure to reapportion despite widespread population shifts that had made urban districts vastly more populous than their rural counterparts. The ramifications of the case were clearly national, because urban and especially suburban Americans were significantly underrepresented in state legislatures throughout the country. Thus, although the Court limited itself to declaring that state legislative reapportionment was justiciable, leaving more specific action in the case to the lower courts, its decision was immediately seen as a revolutionary step – one the Court had repeatedly declined to take.
The immediate consequence of Baker was more litigation. Indeed, within a year of the decision, all but 14 states were involved in reapportionment suits, and the Supreme Court used some of these cases to stake out a clearer substantive position.
Previous research has shown that the presence of behavioral and psychological symptoms (BPSD) can increase the burden of care in dementia (e.g. Donaldson et al., 1998). Impaired executive functions, also called “dysexecutive syndrome” (Wilson et al., 1996), are frequent in dementia and can adversely affect social behavior in everyday life. Reduced awareness of cognitive deficits is associated with impaired executive functions, and may play a significant role in the onset and maintenance of BPSD. A recent paper by Davis and Tremont (2007) demonstrated that dysfunction of frontal systems is associated with increased burden of care in dementia, even after controlling for clinical severity and caregiver mood. In this letter, we report results that support this finding.
In this chapter, we review the line of argument we have pursued in this part of the book. First, we contrast our view of the increasing incumbency advantage and the vanishing marginals with previously dominant notions. Second, we argue that the question to ask is at least as much about improved electoral coordination as it is about any increase in the “real” incumbency advantage. Third, we review our explanation of why electoral coordination improved and briefly consider some possible alternatives. Fourth, we review our explanation of why the two parties' patterns of vote loss in open seats were so different, both across time and cross-sectionally. Fifth, we reconsider the normative stakes involved in the increasing incumbency advantage.
the incumbency advantage, the vanishing marginals, and democratic dominance of the house
The 1960s brought a remarkable complex of changes to House elections: a sudden increase in the incumbency advantage, a sudden decrease in the number of marginal districts, and a sudden eradication of pro-Republican bias outside the South. The dominant way of viewing these changes is still probably that suggested in the work of Erikson (1972) and Mayhew (1974). In (one reconstruction of) their view, the causal sequence was as follows. (1) The incumbency advantage in House elections – a vote premium that accrued to incumbents per se – increased.
The Supreme Court's reapportionment decisions, beginning with Baker v. Carr in 1962, had far more than jurisprudential consequences. They sparked a massive wave of extraordinary redistricting in the mid-1960s. Both state legislative and congressional districts were redrawn more comprehensively - by far - than at any previous time in America's history. Moreover, they changed what would happen at law should a state government fail to enact a new districting plan when one was legally required. This book provides a detailed analysis of how judicial partisanship affected redistricting outcomes in the 1960s, arguing that the reapportionment revolution led indirectly to three fundamental changes in the nature of congressional elections: the abrupt eradication of a 6% pro-Republican bias in the translation of congressional votes into seats outside the south; the abrupt increase in the apparent advantage of incumbents; and the abrupt alteration of the two parties' success in congressional recruitment and elections.
The word gerrymander describes a distinctively (albeit not uniquely) American practice, that of redrawing district lines to achieve partisan (or other) advantage. The word also has a distinctively American etymology, dating back to Elbridge Gerry's term as governor of Massachusetts (1810–1812), when political observers made sport of a district drawn by his party that looked something like a salamander.
At the broadest level, indicated by its title, this book is about gerrymandering. The principles of our analysis could be applied to the original Gerry-mander or to any of its various and long line of descendants (for one such effort, see Engstrom 2001).
At a narrower and more specific level, indicated by its subtitle, this book concerns what was arguably the most important change in the practice of American gerrymandering since its invention. Whereas previously the game of drawing salamanders with district lines was limited to legislators and governors, the courts standing scrupulously aside, after 1964 the rules changed. A new process emerged, with new strategic consequences and nuances. We examine how these procedural changes help explain two of the biggest stories in congressional elections since the 1960s: the seemingly invulnerable Democratic majority in the House of Representatives before 1994 and the seemingly unfair and bloated advantage of incumbents over challengers.
the reapportionment revolution
The Supreme Court's reapportionment decisions, beginning with Baker v. Carr in 1962, were soon hailed by legal scholars as revolutionary (see, e.g., Baker 1966, p. 3; Dixon 1968, p. 99). They reversed decades of court decisions that had consistently held that the drawing of legislative district lines, fraught though it was with malapportionment and gerrymandering, was not justiciable.
In this chapter, we develop a general model of the redistricting process in the United States. We use this model to examine two features of how congressional votes translate into seats: partisan bias (how much larger or smaller a party's seat share is than its vote share would warrant) and responsiveness (how much party seat shares respond to changes in vote shares). Bias and responsiveness are standard concepts in the analysis of redistricting and will be defined more fully later. From our model, we derive specific hypotheses about how the bias and responsiveness of a redistricting plan will differ as a function of two conditions obtaining when that plan is enacted: (1) the legally defined reversionary outcome of the redistricting process and (2) which party controls the legislative branches (house, senate, and governorship) of the state. We test our hypotheses in the next chapter.
a model of congressional redistricting
The literature on redistricting categorizes gerrymanders according to the varying goals that those who redraw district lines pursue. For present purposes, the most important categories are proincumbent gerrymanders (when a bipartisan alliance draws the lines to preserve the current incumbents' chances of victory) and partisan gerrymanders (when a single party draws the lines to maximize its seat share). In this section, we develop a model in which these two types of gerrymander – along with a third “mixed” type – emerge endogenously as a function of partisan control of the redistricting process and the nature of the reversionary outcome.
The basic elements of our model are straightforward. There are two parties, the Republicans and the Democrats, in a given state.
In the previous chapter, we showed that conventional estimators overestimate the incumbency advantage. The positive correlation between the incumbent party's vote share and the presence of an incumbent candidate seeking reelection is due not only to the strength of the campaign that the incumbent candidate can mount (using the resources of her office) but also to the fact that incumbents who forecast sufficiently poor vote shares for their party, relative to their opportunity costs, withdraw from the race. Our findings also showed that the contribution of vote forecasting to the incumbency advantage as usually measured increased in the 1960s.
In this chapter, we seek to explain why incumbents' exits might have become more reliable signals that their party would fare poorly in the ensuing open-seat election. We note that the reapportionment revolution introduced unavoidable redistricting after every decennial census that candidates could more easily anticipate. This, in turn, meant that the next redistricting increasingly provided a focal point in the entry game between strong would-be challengers and sitting incumbents.
how anticipations of redistricting help coordinate entry
The Supreme Court's decision in Wesberry v. Sanders should have abruptly changed politicians' expectations regarding the frequency and inevitability of redistricting. Prior to Wesberry, redistrictings were virtually certain only in states that lost seats in the decennial reapportionment. In the other states, those that either held even or gained in the reapportionment, district lines were often preserved. After Wesberry, in contrast, everyone knew that redistrictings would occur at least once every 10 years. Moreover, virtually every district would be affected because population shifts over the decade would leave almost no districts close enough to the new statewide average to satisfy the court's equalpopulation mandate.
Elbridge Gerry was governor of Massachusetts from 1810 to 1812. During his term, his party produced an artful electoral map intended to maximize the number of seats it could eke out of its expected vote share. Contemporary observers latched onto one district in particular, in the shape of a salamander, and pronounced it a Gerry-mander.
This book is about a unique episode in the long history of American gerrymandering – the Supreme Court's landmark reapportionment decisions in the early 1960s and their electoral consequences. The dramatis personae of our story are the state politicians who drew congressional district lines, the judges on the courts supervising their handiwork, and the candidates competing for congressional office. The plot of our story concerns the strategic adaptation of these actors to the new electoral playing field created by the Court's decisions.
In writing our story, we have incurred numerous debts. Here we thank some of those we should, with apologies to those we have inadvertently forgotten. For conversations about and comments on our project, we thank R. Michael Alvarez, Bruce Cain, Andrea Campbell, Chris Den Hartog, Andrew Gelman, Dave Grether, Bernie Grofman, John Mark Hansen, Simon Jackman, Gary Jacobson, Sam Kernell, D. Roderick Kiewiet, Gary King, Morgan Kousser, Mat McCubbins, Mike McDonald, Jonathan Nagler, Steven Smith, Matthew Spitzer, Simon Wilke, and participants in seminars given at Princeton, UC Berkeley, UCLA, UC Riverside, the Hoover Institution, Northwestern University, the University of Minnesota, Yale University, the University of Chicago, the University of Illinois at Urbana-Champaign, Harvard University, and the University of Rochester. For research assistance, we thank Chris Den Hartog, Mike McDonald, Meredith Rolfe, and the Reference Law Librarians at the University of Chicago.
I’m [not running for reelection]. It reflects a lack of confidence on my part that
we’ll be a majority party in the House any time soon.
Rod Chandler (R-Washington) (quoted in Ansolabehere and Gerber 1997)
In Chapter 8, we showed that the incumbency advantage (as measured by the standard methodology) tended to be larger for Republicans than for Democrats after the mid-1960s. In this chapter, we seek to explain this and other party differences.
In outline, our argument is as follows. First, the abrupt disappearance of pro-Republican bias in the translation of congressional votes into congressional seats (outside the South) led to an abrupt structural decline in the short- and medium-term probability that the Republicans would be able to secure a majority in the U.S. House of Representatives. This decline in the Republicans' chances of securing a majority raised the expected value of a House seat for Democrats but depressed it for Republicans, leading to a syndrome of entry and selection effects that, among other things, explain the party differences uncovered in Chapter 8.
The rest of the chapter proceeds as follows. The first section outlines a simple model of a candidate's decision to enter a House race. We then use that model to explain why candidates of the two parties should have had different valuations of a House seat and how the valuation gap evolved over the post-World War II era. The key variable in this discussion is the probability that each party would secure a majority in the House. The third section explains the consequences that should have followed as the valuation gap rose and fell. We then test these predictions, discuss our findings, and conclude.
the decision to enter
Consider a candidate deciding whether to enter a race for a House seat.
Previous studies of the incumbency advantage have focused on how much running an incumbent for reelection boosts the incumbent party's expected vote share. Incumbents, however, can decide whether to run for reelection, and their decisions are based partly on anticipation of the vote share they might win were they to run. Thus, while the presence of an incumbent may boost the incumbent party's vote share, forecasts of this vote share may determine whether there is an incumbent in the race to begin with.
All current measures of the incumbency advantage risk overestimating how much incumbency matters, by neglecting the possibility that incumbents tend to seek reelection when the prospects for their party are better, while retiring when those prospects are poorer. To the extent that incumbents are good at forecasting votes, one will find the incumbent party's vote share larger when there is an incumbent (who correctly forecast the favorable vote and hence ran for reelection) and smaller when there is no incumbent (the incumbent having retired in the face of a bad expected vote, which the nonincumbent replacing her to some extent inherits). In technical terms, the argument just given amounts to saying that there is a species of simultaneity bias afflicting current measures of the incumbency advantage.
It is not just incumbents' entry strategies that pose analytic challenges, however. When high-quality challengers (defined as those who have previously won elective office) enter the fray, they presumably do so partly on the basis of favorable vote forecasts. Their decisions to enter accordingly may bias estimates of how much the presence of a strong challenger boosts the challenging party's vote share.
In the previous chapter, we focused on redistricting plans drawn prior to the courts' entry into the redistricting process. During this precourt period, reversionary plans – that is, the plans that would stand in force should the state legislature and governor be unable to agree on a new plan – were automatic. Both their content and the conditions under which they would come into force were prespecified and did not depend on the decisions of any strategic actor. In contrast, after the Supreme Court's reapportionment decisions, individual courts decided both the content of the reversionary plan and, to some degree, the conditions under which it would be invoked. In this chapter, we begin to consider the political consequences of this change from automatic to discretionary reversions.
The first two sections consider how the partisan complexion of the courts supervising (or potentially supervising) redistricting actions in the 1960s might in theory have affected the plans ultimately used. Based on an extension of the model developed in the previous chapter, we argue that one should expect court partisanship to affect the partisan bias and responsiveness embodied in the implemented plan, even if the court did not impose a plan and even though bias and responsiveness were not justiciable.
The third section provides qualitative evidence that each court's partisan complexion did affect its decisions. The fourth section provides quantitative evidence that each court's partisan complexion, relative to that of the plan it was judging, affected the level of malapportionment it was prepared to tolerate. The final section concludes.