What Happened in Pennsylvania
This is the second part in a multi-part series where I analyze the 2020 presidential election results. In the first part, I looked at Michigan and for those that saw that, you may notice some similar trends playing out in Pennsylvania. I also talked extensively about Pennsylvania before the election, where I identified the various subsections and opportunities for Joe Biden to do well in this state. If you haven’t already, I recommend you check out that article, as I talk more in-depth about the different subsections, their previous voting behavior, and their demographic breakdowns.
With all that said, let’s dive in.
Big Picture
Joe Biden- 3,458,229 (50.01 percent, up 2.55 percent from Clinton in 2016)
Donald Trump- 3,377,674 (48.84 percent, up 0.66 percent from 2016)
FiveThirtyEight Projection- 52 for Biden (up 1.99), 47.3 for Trump (down 1.54)
RCP Average- 48.7 for Biden (down 1.31), 47.5 for Trump (down 1.34)
OurProgress Projection- 50.96 for Biden (up 0.95), 46.39 for Trump (down 2.45)
Here, we have yet another state that Donald Trump flipped in 2016, breaking down the touted “Blue Wall” that Democrats had held in the Rust Belt in the previous two decades. Considering this, Pennsylvania was another “must win” state for the Democratic nominee. Compared to other Rust Belt states, Pennsylvania was widely seen as the most significant in the electoral math. FiveThirtyEight projected that Pennsylvania had the highest probability of becoming the “tipping point” of the election, meaning that if you added up each state’s electoral votes starting with those decided by the largest margin and ending with those decided by the narrowest margin, this state’s electoral votes would most likely put the winner above the 270 vote threshold needed to win. But for Joe Biden, winning Pennsylvania was more than a strategic necessity. This was his birth state and he has spent most of his political career representing Delaware, which carries strong roots to the neighboring Philadelphia metropolitan area. For Biden, this battle was personal.
After multiple days of counting and legal action from the Trump campaign, Joe Biden emerged victorious in Pennsylvania. While it was narrow, Biden’s margin of victory (80,555 votes) was almost double Trump’s margin of victory in 2016 (44,292 votes). Interestingly, Pennsylvania did not end up being the tipping point in the election. Using the counting method stated above, Pennsylvania only got Biden to 269 electoral votes; it was Wisconsin, the next state up with a smaller margin, that pushed him over the finish line. Even so, Pennsylvania was another testament to the success of Biden’s commitment to recapturing the Rust Belt.
As of Election Day, Pennsylvania has had 200,074 COVID cases, ranking 13th in the nation (just above Michigan, in fact). But similar to Michigan, Pennsylvania’s large population lowers the incidence rate to just 1.7 percent, below the national average and ranking 43rd among the states. On the other hand, there have been 8,842 deaths, ranking 8th highest in the nation. And the fatality rate, 4 percent, is above the national average and ranks 7th among the states (once again, just above Michigan).
While Pennsylvania has followed a similar trajectory to Michigan on the medical side of the pandemic, the economic hit has been more modest. While nothing to dismiss, Pennsylvania went from a 5.9 percent unemployment rate in March to 15.6 percent in April, a 9.7-point increase, which ranks 22nd among the states.
Considering that the state is home to two major metropolitan areas, Pennsylvania was a major hub of protest activity in recent months. According to the US Crisis Monitor managed by ACLED, there were 703 BLM protests in Pennsylvania between Memorial Day and Election Day, the fourth most of any state. This makes sense, given that it’s the fifth largest state by population, but the demographics most likely to organize and participate in these events are not dominant within that population. Pennsylvania has the 10th highest black population in raw numbers (21st by percentage of the total population), 6th highest amount of college graduates (but 24th by percentage), and the 6th highest young adult population (but 43rd by percentage). So while these demographics aren’t large enough to dominate within the state’s population, they exist in large enough numbers to organize a large number of events.
As for the exit polls, Biden retained 95 percent of Pennsylvania’s Clinton voters while picking up 63 percent of those that did not vote in 2016, compared to Trump only retaining 92 percent of his voters from four years ago.
For race, President Trump won 57 percent of white voters, which is fairly close to his performance nationwide. Meanwhile, Biden only won 42 percent of white voters in Pennsylvania, which is only a 2-point improvement from Clinton in 2016. As for Hispanic voters, Biden won them decisively with 69 percent; however, President Trump made a 5-point improvement among these voters compared to 2016, receiving 27 percent. And similar to Michigan, the black vote in Pennsylvania is almost identical to 2016, where Biden won 92 percent of these voters.
Regarding educational attainment, Biden won 57 percent of college graduates and 54 percent of white college graduates, slightly better than his national performance and a shift from 2016, where Trump was more competitive with these voters. Meanwhile, President Trump won 54 percent of voters without a college degree, including 66 percent of white voters without a college degree. This widens the education gap seen in 2016.
As for age, there was a clear generation gap in Pennsylvania, similar to that seen nationwide. Biden was the clear favorite among younger voters, winning 62 percent of voters in the 18–29 age bracket and 60 percent of those in the 30–44 age bracket (marking a 10-point and 7-point improvement from Clinton respectively). On the other hand, President Trump won 58 percent of voters in the 45–64 age bracket and 53 percent of those in the 65 and over bracket (a 6-point improvement and a 1-point loss from his 2016 performance). Compared to the rest of the nation, this divide is more cleanly defined, as Biden won the 30–44 bracket and Trump won the 45–64 bracket by more decisive margins than their national numbers in these groups. As a result, it really is more of a gap in Pennsylvania between the under 45 voters and the over 45 voters, rather a spectrum as seen nationwide.
Geographically, Pennsylvania is more starkly divided than the rest of the nation. On one hand, Biden won 74 percent of its urban voters, compared to just 60 percent nationwide and a 4-point improvement from Clinton in 2016. But on the other hand, Trump won 69 percent of its rural voters, compared to just 57 percent nationwide and a 2-point loss from 2016. Meanwhile, suburban voters were closely contested, similar to the rest of the nation; however in Pennsylvania, President Trump had the slight edge with 51 percent, just a 1-point loss from 2016.
Socioeconomically, only 10 percent of Pennsylvania voters said that the pandemic has caused them severe financial hardship (compared to 16 percent nationwide), of which 76 percent backed Joe Biden. 38 percent said that the pandemic has caused them moderate financial hardship (compared to 39 percent nationwide), of which 53 percent supported Biden. And 51 percent said that the pandemic has caused them no financial hardship (compared to just 44 percent nationwide), of which 55 percent supported President Trump. Regarding income brackets, the results were strange compared to the rest of the nation. Biden won 56 percent of voters making under $50,000 (close to the 55 percent he received nationally); however, President Trump edged him out by 2 points among voters in the $50,000-$99,999 bracket, winning 51 percent (compared to Biden’s 15-point rout nationwide). But among voters making over $100,000, President Trump only won by 4 points (compared to his win by 12 points nationwide). Biden made sizable improvements from Clinton across all three brackets; however, Biden’s lack of competitiveness among voters in the middle bracket is unusual compared to national trends.
When asked about their most important issue, 38 percent of Pennsylvania voters said the economy (compared to 35 percent nationwide), of which 85 percent voted for President Trump (compared to 83 percent nationwide). Meanwhile, 20 percent said the pandemic was the most important issue and 14 percent said racial inequality, of which 91 percent and 92 percent respectively backed Joe Biden. When asked what was more important to accomplish, 53 percent of Pennsylvania voters prioritized containing the virus (of which, 85 percent voted for Biden) and 43 percent prioritized reopening the economy (of which, 88 percent voted for President Trump). This divide is consistent with national trends.
Now that we’ve gone over the exit polls, let’s run some regressions. First, here are various regression results for Biden’s county-level vote share.
Similar to the regression results found in Michigan, key developments in 2020, such as the pandemic, the rise in unemployment, and uptick in BLM protest activity did not have much of an impact on Biden’s vote share in Pennsylvania. With that said, we do see statistically significant relationships with different demographic groups. Race and ethnicity are key examples, as both the share of black residents and the share of Hispanic residents are positively associated with a county’s Biden vote share. The education gap, which was a strong and consistent predictor of Biden’s vote share in Michigan, was also powerful in Pennsylvania, where Biden was more successful in more highly educated counties. Other than that, socioeconomic indicators were statistically significant, where Biden was more successful in more affluent counties as well as counties with higher income inequality. These results are similar to those in Michigan.
Next, let’s look at the regression models for the Democratic vote share change between 2016 and 2020.
Here, we also see similar results to that of Michigan. For the most part, none of the pandemic or protest variables are statistically significant, indicating that the shocks induced by these developments did not change Democratic vote share. Despite this, there is a negative relationship for the September unemployment rate (B= -0.36; -0.35) that is statistically significant in Models 1 and 3. This provides some evidence that in areas where unemployment remained high even after the summer recovery, Biden underperformed from Clinton in 2016. No such relationship was observed in Michigan, which took a much larger spike in unemployment in April than Pennsylvania. Even so, this finding suggests that lingering economic problems did not help Joe Biden.
Outside of that, the education gap widened between 2016 and 2020, as Biden did better than Clinton in highly educated counties while President Trump improved in counties with lower high school completion rates. There’s actually some evidence that Biden underperformed Clinton in younger counties, although his overall support with young adults remained very high. And there was also a slight improvement for Biden in more affluent counties.
Now that we’ve described some broad trends within the state, let’s break down the results by subsection.
Solidly Democratic Counties
Allegheny- Biden Hold
Biden- 429,065 (59.61 percent, up 3.07), Trump- 282,324 (39.23 percent, down 0.69)
OurProgress Projection- 59.67 for Biden (up 0.06 from actual), 40.33 for Trump (up 1.1)
Delaware- Biden Hold
Biden- 206,423 (62.95 percent, up 3.34), Trump- 118,532 (36.15 percent, down 1.04)
OurProgress Projection- 65.55 for Biden (up 2.6), 34.45 for Trump (down 1.7)
Montgomery- Biden Hold
Biden- 319,511 (62.63 percent, up 3.72), Trump- 185,460 (36.35 percent, down 1.08)
OurProgress Projection- 65.32 for Biden (up 2.69), 34.68 for Trump (down 1.67)
Philadelphia- Biden Hold
Biden- 603,790 (81.44 percent, down 1.09), Trump- 132,740 (17.90 percent, up 2.54)
OurProgress Projection- 85.74 for Biden (up 4.3), 14.26 for Trump (down 3.64)
First, there are four solidly Democratic counties. As expected, Biden established a solid foundation through his performance here. In this subsection, Biden received 67.8 percent of the vote, a slight improvement from the 66.3 percent Clinton received in 2016. This translates to a net gain of 173,663 votes.
Now interestingly, Biden underperformed in Philadelphia County compared to Clinton four years ago. But considering that he still received 81.4 percent of the vote in Pennsylvania’s largest city, there wouldn’t be much gain from the Biden campaign to get a few extra percentage points there. Not to mention that this minor loss is offset by larger gains elsewhere. Particularly, Biden improved by 3 percent in Allegheny County (Pittsburgh), another large pot of voters. Biden also improved by 3 percent each in Delaware and Montgomery, two sizable suburban counties in the Philadelphia metropolitan area (and fairly close to Biden’s home state of Delaware). So on the whole, Biden benefited a lot from anti-Trump sentiment, higher voter turnout, and his roots in the Philadelphia metro.
Similar to the rest of the country, the pandemic has swept through this subsection. As of Election Day, there have been 91,151 COVID cases in this subsection; however its large population sets the incidence at only 2.2 percent (below the national average). On the other hand, there have been 3,969 deaths, setting the fatality rate at 4.4 percent. About half of the total cases and deaths come from Philadelphia County, which makes sense, given that it’s the largest and most densely populated county. Regarding timing, a plurality of the cases came in the early months. But even as the case increases were smaller in the subsequent months leading up to the election, there were still a substantial number of cases, so the “downturn” wasn’t that significant. And for unemployment, this subsection largely reflects the rest of the state as it went from 5.8 percent in March to 15.6 percent in April. And while it has largely recovered in the months since then, the unemployment rate was still 9 percent as of September (above the national average).
Finally, as of Election Day, there have been 268 BLM protests in this subsection, with 121 in Philadelphia County and 107 in Allegheny. This makes sense, given that the subsection is 25.4 percent black (double the Pennsylvania average), 14.4 percent young adult (above the state average), and three of the four counties have more college graduates than the state average. What’s interesting is that Allegheny, which is only 13.4 percent black, has almost the same number of protests as Philadelphia, which is 43.6 percent black, although this discrepancy can be explained by both of their large populations and the fact that Allegheny has more college graduates than Philadelphia.
Swing Counties
Bucks- Biden Hold
Biden- 204,712 (51.66 percent, up 3.14), Trump- 187,367 (47.29 percent, down 0.45)
OurProgress Projection- 52.91 for Biden (up 1.25), 47.09 for Trump (down 0.2)
Centre- Biden Hold
Biden- 40,055 (51.69 percent, up 2.98), Trump- 36,372 (46.94 percent, up 0.61)
OurProgress Projection- 54.88 for Biden (up 3.19), 45.12 for Trump (down 1.82)
Chester- Biden Hold
Biden- 182,372 (57.99 percent, up 5.28), Trump- 128,565 (40.88 percent, down 2.32)
OurProgress Projection- 59.91 for Biden (up 1.92), 40.09 for Trump (down 0.79)
Dauphin- Biden Hold
Biden- 78,983 (53.60 percent, up 4.15), Trump- 66,408 (45.06 percent, down 1.44)
OurProgress Projection- 51.95 for Biden (down 1.65), 48.05 for Trump (up 2.99)
Erie- Biden Flip
Biden- 68,286 (49.81 percent, up 2.83), Trump- 66,869 (48.78 percent, up 0.21)
OurProgress Projection- 53.43 for Biden (up 3.62), 46.57 for Trump (down 2.21)
Lackawanna- Biden Hold
Biden- 61,991 (53.71 percent, up 3.47), Trump- 52,334 (45.35 percent, down 1.42)
OurProgress Projection- 57.15 for Biden (up 3.44), 42.85 for Trump (down 2.5)
Lehigh- Biden Hold
Biden- 98,288 (53.21 percent, up 2.70), Trump- 84,259 (45.62 percent, down 0.16)
OurProgress Projection- 55.2 for Biden (up 1.99), 44.8 for Trump (down 0.82)
Monroe- Biden Hold
Biden- 44,060 (52.56 percent, up 3.93), Trump- 38,726 (46.20 percent, down 1.67)
OurProgress Projection- 51.92 for Biden (down 0.64), 48.08 for Trump (up 1.88)
Northampton- Biden Flip
Biden- 85,087 (49.78 percent, up 3.60), Trump- 83,854 (49.05 percent, down 0.93)
OurProgress Projection- 49.1 for Biden (down 0.68), 50.9 for Trump (up 1.85)
Biden also improved from Clinton in this subsection. While Clinton won seven of these nine counties four years ago, Biden won all nine counties this time around with an average per-county vote share increase of 3.6 percent. Ultimately, Biden received 53.1 percent of the vote in this subsection, a 3.7-point improvement from the 49.4 percent Clinton received in 2016 and a 0.6-point improvement from President Obama’s 2012 performance. This translates to a vote total of 863,834 votes, a net gain of 161,689 votes from Clinton in 2016.
While the improvements from 2016 are pretty notable, the map in this subsection is still very different from 2012, the most recent election before Donald Trump entered the scene. Biden only outperformed Obama’s 2012 showing in four of these counties (Bucks, Centre, Chester, and Dauphin), leaving five counties where Biden underperformed. A deeper look at this cleavage reveals an interesting trend. The four counties where Biden did better than Obama in 2012 all had a lower April unemployment rate (13.3 percent), lower September unemployment rate (6.4 percent), a higher median household income (average of $76,000), lower average poverty rate (10.9 percent), and considerably more college graduates than the five counties where Biden underperformed Obama. This is consistent with the findings demonstrated in Michigan and throughout this part, where despite Biden recapturing portions of the “Blue Wall”, President Trump retained considerable support from the working class regions within that Wall, where voters resonated with his framing of public health restrictions as economically destructive.
On the other hand, some of these counties have changed dramatically regardless of the starting point. Chester, a suburban county west of Philadelphia, is the prominent example. For decades, this was a traditionally Republican county. And while it swung to Obama in 2008, it narrowly flipped back to Mitt Romney in 2012 by backing Mitt Romney. But while this seemed like a return to form, it turned out to be a detour in a more lasting transformation. In 2016, Chester went to Clinton with 51.9 percent of the vote, a 2.7-point swing. And this year, it swung an additional 5.3 points, allowing Biden to win it with 57.8 percent. In other words, this suburban and historically Republican county has been transformed into a solidly Democratic county in just eight years. Places like Chester also demonstrate how the two parties have changed their coalitions in a Trumpified political environment, where Democrats have become more successful in the suburbs.
Indeed, the pandemic’s impact on this subsection has been largely economic, but that shouldn’t dismiss the biological damage as well. As of Election Day, there have been 49,661 COVID cases in this subsection, setting the incidence rate at 1.7 percent (below the national average). There isn’t much of a divide here between the counties Biden overperformed Obama versus those where he underperformed. On the other hand, there have been 2,299 deaths, setting the fatality rate at 4.6 percent (above the national average). Regarding timing, the virus hit this subsection pretty hard in the early months. And while it dipped down a bit in the summer months, the virus has since made a resurgence in the early fall months. And for unemployment, the subsection took a 9.6-point hit, going from 5.4 percent in March to about 15 percent in April. As noted above, the counties where Biden underperformed from President Obama took the brunt of this hit. In the months since, the unemployment rate has mostly recovered; however, it still stood at 7.2 percent as of September.
Finally, as of Election Day, there have been 167 BLM protests in this subsection. 45 have been held in the aforementioned Chester County, while an additional 29 have occurred in Erie (which will be discussed more in the next section), and 24 in Bucks. Admittedly, this subsection is not very diverse racially, as only 8.1 percent of its population is black (below the Pennsylvania average); however, Chester is the most diverse, with 19.2 percent of its population being black. Young adults account for 12.9 percent of the population (on par with the state average), but there are some hotspots, such as Centre County, home to Penn State University. But the biggest asset for this subsection is the large number of college graduates, especially in the four counties where Biden outperformed President Obama. In three counties, over 40 percent of the adult population holds a bachelor’s degree and in Centre County, that figure is over 50 percent. On the whole, the high concentration of college graduates in the subsection along with key areas with high concentrations of young adults and black residents increased protest activity.
Obama-Trump Counties
Erie- Biden Flip
Biden- 68,286 (49.81 percent, up 2.83), Trump- 66,869 (48.78 percent, up 0.21)
OurProgress Projection- 53.43 for Biden (up 3.62), 46.57 for Trump (down 2.21)
Luzerne- Trump Hold
Biden- 64,873 (42.31 percent, up 3.45), Trump- 86,929 (56.70 percent, down 1.60)
OurProgress Projection- 46.37 for Biden (up 4.06), 53.63 for Trump (down 3.07)
Northampton- Biden Flip
Biden- 85,087 (49.78 percent, up 3.60), Trump- 83,854 (49.05 percent, down 0.93)
OurProgress Projection- 49.1 for Biden (down 0.68), 50.9 for Trump (up 1.85)
Finally, of the three Obama-Trump counties, Biden won back two of them (both of which overlap with the swing counties). On average, he increased his vote share in each county by an average of 3.3 percent. Overall, Biden received 47.3 percent of the vote which, while up from the 44 percent Clinton received in 2016, is still below the benchmark of 53.5 percent that President Obama received in 2012. In terms of raw votes, this translates to a net gain of 41,411 votes from Clinton in 2016.
Across the board, the misses on the OurProgress projection weren’t as big in Pennsylvania as they were in Michigan (or some of the other states that will be discussed later), but they were still prevalent. For this subsection, the projection model underestimated Biden’s performance in Northampton, which was expected to narrowly favor President Trump. But in the end, Biden made enough pushes to flip back this county.
On the other hand, Biden was overestimated in Luzerne, a county that was already expected to remain in Trump’s corner. I already discussed Luzerne in my pre-election article, particularly how it has one of the fastest growing Hispanic populations in the United States and how it had a massive swing towards Trump in 2016. This time around, Biden did improve Clinton by about 3 percent, but this county remained in the Trump column this time around. Based on these results, there’s evidence that while the large Hispanic population helped Biden somewhat, these gains weren’t enough to offset those made by Trump four years ago.
As of Election Day, there have been 13,779 COVID cases in this subsection, setting the incidence rate 1.5 percent (below the national average). On the other hand, there have been 564 deaths due to the virus, setting the fatality rate at 4.1 percent (above the national average). Regarding timing, about half of the cases came in the early months of the pandemic, indicating an early hit. But while the cases dipped down significantly during the summer months, there has been a resurgence in the early fall months. And for unemployment, this subsection took a 10.4-point hit, going from 6.7 percent in March to 17.1 percent in April. And while there has been a recovery for the most part, unemployment was still at 8.3 percent as of September. It should be noted that for all three months, the unemployment rate was the highest in Luzerne, the one county that President Trump retained in this year’s election.
Finally, as of Election Day, there have only been 69 BLM protests. Interestingly, 29 of them occurred in Erie County while 25 occurred in Luzerne. And considering how these two counties diverged quite a bit in their voting behavior, the fact that they each experienced comparable protest activity suggests that the protests alone did not cause a significant effect on their respective vote shares. Demographically, the subsection is only 7.2 percent black (below the state average) and 12.7 percent young adult (on par with the state average), while this distribution is roughly equal in each county. Furthermore, these counties are below the state average for college graduation, where no more than 30 percent of adults hold a bachelor’s degree in any of them. Overall, protest activity is noticeably lower in this subsection than in others discussed in this article due to their unfavorable demographics and less progressive political leanings.
Reclassifications
Similar to Michigan, there are a few reclassifications for counties based on how their voting behavior changed from 2016 to 2020. These changes are reflected below.
Berks (Solidly Republican to Swing)
Chester (Swing to Solidly Democratic)
Erie (Obama-Trump to Trump-Biden)
Northampton (Obama-Trump to Trump-Biden)
Admittedly, there are fewer reclassifications than some states that will be discussed in later parts. But considering that on average the Democratic vote share in each county only increased by 2.1 percentage points, the 2020 map didn’t change too drastically from the 2016 map. Most of the improvements Biden made were in counties that were either solidly Democratic already or were swing counties that leaned Democratic.
Conclusion
In my prediction article, I noted how Pennsylvania would be a close state based on the polling; however, I had confidence that Biden would pick up ground in the swing counties and the Obama-Trump counties. And I feel like to a large extent, it lived up to my expectations. It was the fourth closest state in the 2020 election, being decided by just 1.16 points, and Biden flipped two of the three Obama-Trump counties. And while the OurProgress projection and other models generally underestimated President Trump, they did a better job at capturing the closeness of the race in Pennsylvania.
But similar to Michigan, while Biden was more competitive than Hillary Clinton in many of the working class areas of Pennsylvania, the bulk of Biden’s success came in major cities, affluent suburbs, and highly educated areas (such as Centre County). And while Biden did flip two Obama-Trump counties (Erie and Northampton), these areas were already more competitive in 2016 than most of Michigan’s Obama-Trump counties. To President Trump’s credit, he did a great job at mitigating his losses in the counties that are less affluent, have fewer college graduates, and have a history of being strong industrial and manufacturing sectors. This speaks to Trump’s ability to convert his opposition to pandemic-induced restrictions and shutdowns into emboldened support in areas that would bear the greatest economic brunt to such policies.
Despite the disparities in economic outcomes across these subsections, there aren’t as many disparities regarding the pandemic’s medical severity. All three subsections have relatively low COVID incidence rates and above average fatality rates; however, the minor differences between them are not substantial enough to explain their differences in voting behavior. This finding is reinforced by the regression results, that fail to find a statistically significant relationship between the Biden vote share and Democratic vote share change with any of the metrics for the pandemic’s severity.
Finally, while the regression results didn’t find a statistically significant relationship, there is a sharp difference in protest activity across the subsections. In the solidly Democratic counties, both Allegheny and Philadelphia were major hubs for BLM protests, given their large and diverse populations. In the swing counties, areas with a lot of college graduates experienced an uptick in protest activity in the months leading up to the election. And the Obama-Trump counties experienced far fewer protests, in part to their less progressive disposition and unfavorable demographics. Considering this evidence, while protests were higher in counties that supported Joe Biden, areas that had a lot of protests also had favorable demographics, such as a high share of college graduates, high share of young adults, or high share of black residents. Similar to Michigan, protest activity isn’t associated with voting behavior so much as it is a proxy for demographic variables which are associated with preferences.
Overall, Pennsylvania was both more competitive and more static than Michigan. For the most part, the race was decided by further consolidating support in solidly Democratic counties while making modest improvements in the swing and Obama-Trump counties. While this doesn’t sound particularly impressive, it allowed Joe Biden to recapture his birth state after it had gone to Trump just four years earlier.
So that’s it for the second part of this series. If you enjoyed this, please like and follow the Book Aisle. Also share this article on Facebook, LinkedIn, and other social media platforms.
